Customary International Law as U.S. Law: A Critique of the

Georgetown University Law CenterScholarship @ GEORGETOWN LAW

2011

Customary International Law as U.S. Law: ACritique of the Revisionist and IntermediatePositions and a Defense of the Modern PositionCarlos Manuel VázquezGeorgetown University Law Center, [email protected]

Georgetown Public Law and Legal Theory Research Paper No. 12-070

This paper can be downloaded free of charge from:http://scholarship.law.georgetown.edu/facpub/972http://ssrn.com/abstract=2083120

This open-access article is brought to you by the Georgetown Law Library. Posted with permission of the author.Follow this and additional works at: http://scholarship.law.georgetown.edu/facpub

Part of the Comparative and Foreign Law Commons, Constitutional Law Commons, Courts Commons, International Law Commons, and theState and Local Government Law Commons

86 Notre Dame L. Rev. 1495-1634 (2011)

http://www.law.georgetown.edu/?utm_source=scholarship.law.georgetown.edu%2Ffacpub%2F972&utm_medium=PDF&utm_campaign=PDFCoverPages



http://scholarship.law.georgetown.edu?utm_source=scholarship.law.georgetown.edu%2Ffacpub%2F972&utm_medium=PDF&utm_campaign=PDFCoverPages

http://scholarship.law.georgetown.edu/facpub?utm_source=scholarship.law.georgetown.edu%2Ffacpub%2F972&utm_medium=PDF&utm_campaign=PDFCoverPages

http://network.bepress.com/hgg/discipline/836?utm_source=scholarship.law.georgetown.edu%2Ffacpub%2F972&utm_medium=PDF&utm_campaign=PDFCoverPages





\\jciprod01\productn\N\NDL\86-4\NDL403.txt unknown Seq: 1 4-OCT-11 12:26

CUSTOMARY INTERNATIONAL LAW AS U.S. LAW:

A CRITIQUE OF THE REVISIONIST AND

INTERMEDIATE POSITIONS AND A

DEFENSE OF THE MODERN POSITION

Carlos M. Vazquez*

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1496 R

I. THE MODERN POSITION: EXPLICATION AND PRELIMINARY

DEFENSE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1515 R

A. The Basic Case for the Modern Position . . . . . . . . . . . . . . . . . . 1516 R

1. Constitutional Structure and Original Intent . . . . . . 1517 R

2. Pre-Erie Doctrine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1522 R

a. State Authority to Depart from CustomaryInternational Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1523 R

b. Supreme Court Review of CustomaryInternational Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1525 R

c. The General Law as an Intermediate Status . . . 1533 R

3. Post-Erie Doctrine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1535 R

B. The Limits of the Modern Position . . . . . . . . . . . . . . . . . . . . . . . 1538 R

1. Sabbatino and the Inapplicability of SomeCustomary International Law Norms to Some Actsof Foreign States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1539 R

2. The Paquete Habana and the Applicability ofCustomary International Law to Federal Officials . 1544 R

C. Sosa and the Modern Position . . . . . . . . . . . . . . . . . . . . . . . . . . 1546 R

2011 Carlos M. Vazquez. Individuals and nonprofit institutions mayreproduce and distribute copies of this Article in any format, at or below cost, foreducational purposes, so long as each copy identifies the authors, provides a citationto the Notre Dame Law Review, and includes this provision and copyright notice.

* Professor of Law, Georgetown University Law Center. I am grateful forcomments from David Bederman, A.J. Bellia, Curtis Bradley, Bradford Clark, SarahCleveland, William Dodge, Chimene Keitner, John Parry, Wendy Collins Perdue,Stephen Vladeck, and participants in the Fourth Annual Potomac Foreign RelationsRoundtable at the George Washington University School of Law. I am also indebtedto Timothy Work and Eric T. Schmitt for research assistance. I dedicate this Article tothe memory of Professor Louis Henkin (1917–2010).

1495


1496 notre dame law review [vol. 86:4

II. THE INTERMEDIATE THEORIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1554 R

A. Ramsey’s Position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1555 R

1. Nonpreemptive Federal Law as State Law . . . . . . . . . 1556 R

2. Ramsey’s Textual and Historical Support . . . . . . . . . 1563 R

B. Young’s Position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1569 R

1. Young’s Criticisms of the Modern Position . . . . . . . . 1570 R

2. Young’s Intermediate Status for CustomaryInternational Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1574 R

a. State Choice-of-Law Rules . . . . . . . . . . . . . . . . . . . . 1575 R

i. The Diversity and Indeterminacy ofExisting Choice-of-Law Approaches . . . . . . . 1576 R

ii. The Inappositeness of Choice-of-Law Rules 1577 R

iii. The Likelihood of Special Choice-of-LawRules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1585 R

iv. The Role of the Federal Courts . . . . . . . . . . . 1586 R

v. Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1588 R

b. Federal Choice-of-Law Rules . . . . . . . . . . . . . . . . . . 1589 R

C. Aleinikoff’s Position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1592 R

D. The Bellia-Clark Position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1597 R

E. The Bradley-Goldsmith-Moore Position . . . . . . . . . . . . . . . . . . . . 1609 R

III. THE MODERN POSITION, REDUX . . . . . . . . . . . . . . . . . . . . . . . . . . . 1617 R

A. The New Ways of Making Customary International Law . . . 1617 R

B. The New Topics Addressed by Customary International Law 1623 R

IV. STATE INCORPORATION OF CUSTOMARY INTERNATIONAL LAW 1625 R

CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1633 R

INTRODUCTION

In a recent referendum, the citizens of Oklahoma overwhelm-ingly approved a State constitutional amendment providing that thecourts of the State “shall not consider international law or Sharia law”in rendering their decisions.1 The amendment’s exclusion of Sharialaw has garnered most of the media attention,2 but more consequent-

1 State Question No. 755, Legislative Referendum No. 355 (amending OKLA.CONST. art. 7, § 1.), available at https://www.sos.ok.gov/documents/questions/755.pdf; see Jess Bravin, Oklahoma Shariah Ban Halted, WALL ST. J., Nov. 9, 2010, at A6;Aaron Fellmeth, International Law and Foreign Laws in the U.S. State Legislatures, ASILINSIGHTS (May 26, 2011), http://www.asil.org/pdfs/insights/insight110526.pdf.

2 See Jess Bravin, Oklahoma Is Sued over Shariah Ban, WALL ST. J., Nov. 5, 2010, atA5; Carla Hinton, Measure Keeps Courts from Considering Sharia Law, OKLAHOMAN, Nov.5, 2010, at 1A; Donna Leinwand, States Enter Debate on Sharia Law, USA TODAY, Dec. 9,2010, at 3A; James C. McKinley, Jr., Oklahoma Surprise: Islam as an Election Issue, N.Y.TIMES, Nov. 15, 2010, at A12; Darla Slipke, English-Only, Health Care, Sharia Law Mea-sures Pass, OKLAHOMAN, Nov. 3, 2010, at 5A. This aspect of the amendment has been


2011] customary international law as u.s . law 1497

ial by far is the measure’s directive to the State courts to disregardinternational law. Similar measures have been proposed in otherStates, some of them merely barring consideration of Sharia law orforeign law,3 but others barring consideration of international law aswell.4 These measures are clearly unconstitutional insofar as theywould prohibit the State courts from enforcing one of the two mainforms of international law—treaties—as the U.S. Constitution by itsterms requires State courts to give effect to the nation’s treaties, “anyThing in the Constitution or Laws of any State to the Contrary not-withstanding.”5 But the federal Constitution does not expresslyaddress the status of the other principal form of international law—customary international law, or the unwritten law that governs therelations among states and “results from a general and consistentpractice of states followed by them from a sense of legal obligation.”6

These proposed State laws thus starkly raise the question whether theStates may prohibit their courts from giving effect to the UnitedStates’ obligations under customary international law.7

The answer provided by the Restatement (Third) of Foreign RelationsLaw is a clear “no.” Reflecting the settled view regarding the status ofcustomary international law in the U.S. legal system at the time that itwas approved in 1987, the Restatement asserts that such law has thestatus of federal law.8 As such, it preempts inconsistent State law;

challenged on Establishment Clause grounds, see Complaint Seeking a TemporaryRestraining Order & Preliminary Injunction at 6, Awad v. Ziriax, 754 F. Supp. 2d 1298(W.D. Okla. 2010), and a federal court has stayed its enforcement, see Awad, 754 F.Supp. 2d at 1308 (issuing a preliminary injunction). An appeal is pending before theTenth Circuit. See Fellmeth, supra note 1, at 3 & 6 n.27. R

3 See Bill Raftery, An Examination of 2011 Sharia Law & International Law BansBefore State Legislatures, GAVEL TO GAVEL (Nat’l Ctr. for State Courts, Williamsburg,Va.), Jan. 27, 2011, http://gaveltogavel.us/site/2011/01/27/an-examination-of-2011-sharia-law-international-law-bans-before-state-legislatures (discussing proposed legisla-tion in Arkansas, Georgia, Indiana, Kansas, Mississippi, South Carolina, and Texas).

4 See id. (discussing proposed constitutional amendments in Wyoming, Arizona,and South Dakota, proposed legislation in Nebraska that would expressly prohibitconsideration of international law, and proposed legislation in Louisiana and Alaskathat would prohibit the consideration of “foreign law,” defining that term in a waythat appears to include international law).

5 U.S. CONST., art. VI, cl. 2.6 RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW, § 102(1)(c)(2) (1987).7 The term “state,” as used in international law, refers to a nation, such as the

United States or France. In this Article, the term will be used in its internationalsense when it is not capitalized. When capitalized, it will refer to a State of the Union,such as New York or California.

8 See RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW § 111(1) & cmt. d(1987).



State courts must follow federal court interpretations of it; and Statecourt interpretations of it are reviewable in the federal courts. A dec-ade later, however, Professors Curtis Bradley and Jack Goldsmith pub-lished a critique of the Restatement view, which they denominated the“modern position.”9 While acknowledging that the modern positionwas “well-entrenched,” that almost every court that had consideredthe question in the previous twenty years had endorsed it, and that theview was widely regarded as “settled,”10 Bradley and Goldsmith arguedthat the modern position should now be rejected because it is basedon a misinterpretation of pre-Erie11 decisions and is inconsistent with“well-accepted notions of American representative democracy, federalcommon law, separation of powers, and federalism,”12 and becausemodern customary international law is problematic in a number ofrespects.13 For most of the nation’s history, they argued, customaryinternational law was regarded as “general common law,” not federallaw.14 After Erie rejected the concept of general common law, custom-ary international law could have the status of domestic law only if itwas given such status by the federal political branches or by theStates.15 Thus, according to the Bradley-Goldsmith critique, theOklahoma amendment would not only validly bar the Oklahomacourts from considering customary international law that has notbeen incorporated as federal law, it would bar the federal courts fromdoing so as well.

In response to the revisionist challenge to the modern position,numerous scholars have defended the Restatement view.16 Other schol-

9 Curtis A. Bradley & Jack L. Goldsmith, Customary International Law as FederalCommon Law: A Critique of the Modern Position, 110 HARV. L. REV. 815 (1997) [hereinaf-ter Bradley & Goldsmith, Critique]; see also Curtis A. Bradley & Jack L. Goldsmith, III,The Current Illegitimacy of International Human Rights Litigation, 66 FORDHAM L. REV. 319(1997) [hereinafter Bradley & Goldsmith, Current Illegitimacy] (expanding upon theircritique of the modern position); Curtis A. Bradley & Jack L. Goldsmith, Commen-tary, Federal Courts and the Incorporation of International Law, 111 HARV. L. REV. 2260(1998) (responding to critics of their revisionist position).

10 See Bradley & Goldsmith, Critique, supra note 9, at 816–17. R11 Erie R.R. Co. v. Tompkins, 304 U.S. 64 (1938).12 Bradley & Goldsmith, Critique, supra note 9, at 821. R13 Id.14 See id. at 850.15 See id. at 863, 868, 870.16 See, e.g., Ryan Goodman & Derek P. Jinks, Filartiga’s Firm Footing: International

Human Rights and Federal Common Law, 66 FORDHAM L. REV. 463 (1997); HaroldHongju Koh, Commentary, Is International Law Really State Law?, 111 HARV. L. REV.1824 (1998); Gerald L. Neuman, Sense and Nonsense About Customary International Law:A Response to Professors Bradley and Goldsmith, 66 FORDHAM L. REV. 371 (1997); BethStephens, The Law of Our Land: Customary International Law as Federal Law After Erie, 66



ars have advanced what they have characterized as intermediate posi-tions, arguing that customary international law is properly understoodto have a status in between federal and State law. Professor MichaelRamsey has argued that customary international law should beregarded as a form of nonpreemptive federal law.17 Professor ErnestYoung, expanding upon an argument by Professor Arthur Weisburd,18

has argued that customary international law should continue to beunderstood as general law.19 He has proposed the use of choice-of-law rules to determine the applicability of customary international lawin our courts. Dean Alexander Aleinikoff has argued that customaryinternational law should be regarded as “nonpreemptive, nonfederallaw” applicable in the federal courts but not the State courts.20

More recently, Professors A.J. Bellia and Bradford Clark haveadvanced a different sort of intermediate proposal.21 Rather thanpropose an intermediate status for customary international law, Belliaand Clark argue that some categories of customary international law pre-empt State law and others do not.22 Their conclusion that State law issometimes preempted by customary international law is particularlynoteworthy because revisionists rely heavily on a textual and structuralargument closely identified with Professor Clark23—the view that theSupremacy Clause’s list of the categories of preemptive federal law is

FORDHAM L. REV. 393 (1997). Other scholars have sided with the revisionists. See, e.g.,Julian G. Ku, Customary International Law in State Courts, 42 VA. J. INT’L L. 265 (2001)[hereinafter Ku, State Courts]; Julian G. Ku, The State of New York Does Exist: How theStates Control Compliance with International Law, 82 N.C. L. REV. 457 (2004); Julian Ku &John Yoo, Beyond Formalism in Foreign Affairs: A Functional Approach to the Alien TortStatute, 2004 SUP. CT. REV. 153.

17 See MICHAEL D. RAMSEY, THE CONSTITUTION’S TEXT IN FOREIGN AFFAIRS 348–55(2007); Michael D. Ramsey, International Law as Non-Preemptive Federal Law, 42 VA. J.INT’L L. 555, 577, 584 (2002).

18 See A.M. Weisburd, State Courts, Federal Courts, and International Cases, 20 YALE J.INT’L L. 1, 29–35, 51 (1995).

19 Ernest A. Young, Sorting Out the Debate over Customary International Law, 42 VA. J.INT’L L. 365, 370 (2002). The idea that customary international law can continue tobe regarded as general law seems to have been endorsed by Judge William A.Fletcher, International Human Rights in American Courts, 93 VA. L. REV. 653, 672 (2007).

20 See T. Alexander Aleinikoff, International Law, Sovereignty, and American Constitu-tionalism: Reflections on the Customary International Law Debate, 98 AM. J. INT’L L. 91,97–100 (2004).

21 See Anthony J. Bellia Jr. & Bradford R. Clark, The Federal Common Law ofNations, 109 COLUM. L. REV. 1, 1 (2009).

22 See id. at 5–6.23 See Bradley & Goldsmith, Critique, supra note 9, at 855–59; Young, supra note R

19, at 413. R



exhaustive.24 Bellia and Clark argue that the Constitution itself—spe-cifically, its allocation of the war and foreign relations powers to thefederal government—implicitly preempts State laws that conflict witha subset of customary international law.25 They find support in nine-teenth-century Supreme Court decisions for the claim that State law ispreempted if it violates the “perfect rights” of foreign sovereignsunder international law. Violation of perfect rights was regardedunder international law as a justification for going to war.26 State vio-lation of such rights thus interfered with the federal government’sexclusive power to declare war and to conduct the nation’s foreignrelations. The preemptive force of (some) customary internationallaw, on this analysis, is an inference from the constitutional struc-ture—specifically, the Constitution’s allocation of power over war andforeign affairs to the federal government.

Finally, Professors Bradley and Goldsmith themselves haveadvanced what might be regarded as an intermediate position. Intheir original critique, they insisted that customary international lawlacked the status of federal law unless incorporated as such by thefederal political branches.27 The implication was that an act of federallawmaking—a federal statute or a treaty—was required. Portions oftheir critique suggested that this was indeed what they contem-plated.28 Elsewhere, however, they stated that, “to some extent, thePresident [has] the authority to incorporate [customary interna-tional] law into domestic law,”29 and they noted but did not endorsethe possibility that the relevant authorization may be inferred from ajurisdictional statute.30 More recently, writing after the Supreme

24 See Bradford R. Clark, Separation of Powers as a Safeguard of Federalism, 79 TEX. L.REV. 1321, 1323–24, 1429 (2001),

25 See Bellia & Clark, supra note 21, at 8. R

26 See id. at 5.27 See Bradley & Goldsmith, Critique, supra note 9, at 868, 870. R

28 See id. at 825–26 (stating that “[t]he Constitution ensured federal control over[customary international law] through two means,” both involving congressionalaction); id. at 858 (“There is . . . no mention of [customary international law] in themenu of supreme federal law in Article VI.”); id. at 868 (relying on “the thesis that theConstitution, by providing for the representation of [S]tate interests in Congress,entrusts the maintenance of the federal balance to ‘the internal safeguards of thepolitical process’” (quoting Garcia v. San Antonio Metro. Transit Auth., 469 U.S. 528,531 (1985)).

29 Id. at 871. But cf. id. at 868 (recognizing that the argument for a presidentialpower to incorporate customary international law is “less forceful[ ]” than for a con-gressional power).

30 See id. at 872–73.



Court in Sosa v. Alvarez-Machain31 endorsed the judicial creation offederal common law rights of action for damages for violations ofsome norms of customary international law,32 they embraced andelaborated on these earlier suggestions. Their post-Sosa article, writ-ten with Professor David Moore, endorses the propriety of readingjurisdictional statutes under certain circumstances as authorization forthe incorporation of customary international law as federal law, andexpands upon their earlier acceptance of a presidential power toincorporate customary international law as federal law.33

This Article offers a critique of the intermediate positions and, inthe process, explicates and defends the modern position. Critics ofthe modern position often describe it as the claim that customaryinternational law has the force of federal law always and for all pur-poses. But this uncompromising conception of the modern positionis a phantom. Adherents of the modern position have always acceptedthat not all of customary international law binds foreign states or thefederal Executive as a matter of U.S. domestic law. The heart of themodern position is that customary international law binds State actorsand thus preempts State law applicable to State officials and privateparties. The basic case for the modern position relies on an inferencefrom the constitutional structure very similar to the one advanced byBellia and Clark: Violations of customary international law risk retalia-tion against the nation as a whole. Permitting States to violate itallows States to externalize the costs of such violations, thus likely pro-ducing excessive violations.

Part I explicates and offers a preliminary defense of the modernposition. It sets forth the affirmative case for the modern positionbased on constitutional structure, original intent, and pre- and post-Erie doctrine, responding to arguments put forward in the initial waveof revisionist scholarship, but deferring to Part II responses to criti-cisms raised by scholars advancing intermediate positions. Part Ishows that the basic structural case for the modern position was wellunderstood by the Founders. Viewed in the light most favorable tothe revisionist view, the evidence of original intent and the pre-Eriecases reflect two contending positions. The first is that the Constitu-tion itself preempts State conduct that violates the state-to-state por-tion of the law of nations. The other is that customary international

31 542 U.S. 692 (2004).32 See id. at 732.33 See Curtis A. Bradley, Jack L. Goldsmith & David H. Moore, Sosa, Customary

International Law, and the Continuing Relevance of Erie, 120 HARV. L. REV. 869, 919–24(2007).



law had the status of general common law. Before Erie, the generalcommon law was understood as different from either federal or Statelaw, but was closer in operation to modern-day federal law than tomodern-day State law. No one claimed that customary internationallaw had a status comparable to modern-day State law.

Part II examines the intermediate positions and concludes thatall but that of Bellia and Clark suffer from fundamental flaws. Ram-sey’s concept of “nonpreemptive federal law” is another name forState law. Thus, Ramsey’s approach would replicate one of theproblems that most concerned the Founders—the lack of federal judi-cial power to prevent or remedy violations of customary internationallaw by the States. Young’s proposal to employ choice-of-law rules todetermine the applicability of customary international law satisfiesErie’s requirement that all law applied in this country’s courts beeither State or federal, but only because choice-of-law rules are them-selves creatures of either State or federal law. To the extent thatYoung would relegate the applicability of customary international lawto State choice-of-law rules, his proposal would present severe difficul-ties stemming from the indeterminacy and inappositeness of suchrules, and, like Ramsey’s approach, would reproduce the problemthat most concerned the Founders.34 Young’s approach would allevi-ate these problems by allowing for the use of federal choice-of-lawrules in some contexts, but he emphasizes that such rules would beapplicable very rarely. Aleinikoff’s approach would violate the oneprinciple that all agree Erie establishes: that the substantive lawapplied in the State and federal courts must be the same. The inter-mediate position of Bradley, Goldsmith, and Moore is problematicbecause it would place inapposite limits on the judiciary’s ability toenforce customary international law as federal law.

The intermediate approach proposed by Bellia and Clark is thor-oughly convincing, but it is not really intermediate. Their structuralargument for according preemptive force to some customary interna-

34 Because the “intermediate” positions of Ramsey and Young turn out to bequite close to the revisionist position, I include those authors among the revisionistswhen I refer to the latter group (except when I distinguish revisionist scholars fromthose who purport to espouse intermediate views). Cf. Ernest A. Young, Sosa and theRetail Incorporation of International Law, 120 HARV. L. REV. F. 28, 28 (2007), http://www.harvardlawreview.org/media/pdf/young.pdf (“I consider myself at least a fellowtraveler [with the revisionists].”); Michael D. Ramsey, Customary International Law inthe Supreme Court, 1901–1945, in INTERNATIONAL LAW IN THE U.S. SUPREME COURT 225,253 (David L. Sloss, Michael D. Ramsey & William S. Dodge eds., 2011) (recognizingthat his position “resembles” the revisionist position and “is open to many of the sameobjections,” including the objection that “it would allow deliberate [S]tate violationsor misinterpretations of customary international law to be unredressed”).



tional law is basically the same as the strongest argument for the mod-ern position. The flaw in their argument is that they do not take it farenough. Their structural argument actually provides substantial sup-port for most of the modern position.

Part III reconsiders the modern position in the light of the revi-sionists’ argument that the customary international law of today dif-fers in important respects from the state-to-state branch of the law ofnations as known to the Founders and as it existed before Erie. Therevisionists’ concerns about the indeterminacy of customary interna-tional law and the loosening of the requirements for recognizing suchlaw have some validity and relevance, but these concerns can be ade-quately addressed by restricting the range of customary norms havingpreemptive force to those that satisfy a heightened standard of clarityand acceptance. The revisionists’ concerns about the new subjectsaddressed by customary international law—in particular, the fact thatsuch law now addresses how a nation treats its own citizens—does notwarrant any additional restriction.

The final Part of the Article addresses a seldom-analyzed aspect ofthe revisionist position—the claim that norms of customary interna-tional law that lack the force of preemptive federal law may be giventhe force of State law through incorporation by State legislatures orcourts. I argue that, for straightforward reasons, the States lack thepower to make norms of customary international law applicable to for-eign states or officials or federal officials. A State’s incorporation ofsuch norms against its own officials or against private parties wouldpose a less obvious structural problem: because customary interna-tional law evolves through the accumulation of state practice andopinio juris, State court decisions regarding the content of such lawcould, in combination with the acts of other States and foreign states,eventually result in the crystallization of norms of customary interna-tional law that the federal government does not support, or the ero-sion of norms that the federal government does support. State courtdecisions regarding the content of customary international law thusinterfere with the federal executive branch’s recognized power tospeak for the United States at the international plane regarding thecontent of such law. This structural problem can be addressed eitherby denying the States the power to incorporate norms of customaryinternational law or by recognizing the Supreme Court’s jurisdictionto review decisions of the State courts regarding the content of cus-tomary international law even when such law is relevant to the caseonly because it has been incorporated as State law. I conclude thatthe latter solution is preferable and that such review would be consis-tent with Article III.



Before proceeding, a few explanations and clarifications are inorder. First, unlike much revisionist scholarship, my analysis will notfocus on litigation under the alien tort provision of the First JudiciaryAct. This provision, currently codified as § 1350 of Title 28, andknown variously as the Alien Tort Claims Act and the Alien Tort Stat-ute, confers jurisdiction on the federal district courts over “any civilaction by an alien for a tort only, committed in violation of the law ofnations or a treaty of the United States.”35 Although it lay dormantfor much of our history, the statute was invigorated in 1980 by theSecond Circuit’s decision in Filartiga v. Pena-Irala,36 which held thatthe provision conferred jurisdiction over a claim brought by aParaguayan citizen against a Paraguayan official based on torture thatoccurred in Paraguay.37 Since then, § 1350 has served as the jurisdic-tional basis for much human rights litigation, most of it alleging viola-tions of customary international law committed abroad by foreignofficials.38

Cases under § 1350 do not shed much light on the status of cus-tomary international law as federal law because a conclusion that cus-tomary international law is federal law is neither necessary norsufficient to resolve the issues that arise in such cases. Most of thesecases seek damages from the individual officials who committed theviolation, and the most contested legal issue in these cases has beenthe source of the right of action for damages. A showing that custom-ary international law is federal law is insufficient to establish a right ofaction in the typical § 1350 case because international law generallydoes not establish a right to recover damages against individuals whoviolate international law. In general, international law applies to theconduct of state actors, and, when violated, it typically places responsi-bility on the state, not the individual state official.39 In some circum-stances, international law does impose individual responsibility, butthis is almost always in the form of criminal responsibility.40 Thus,

35 28 U.S.C. § 1350 (2006).36 630 F.2d 876 (2d Cir. 1980).37 Id. at 878–79, 889.38 See, e.g., Abebe-Jira v. Negewo, 72 F.3d 844 (11th Cir. 1996); Kadic v. Karad_ic,

70 F.3d 232 (2d Cir. 1995); In re Estate of Ferdinand Marcos, Human Rights Litig., 25F.3d 1467 (9th Cir. 1994); Xuncax v. Gramajo, 886 F. Supp. 162 (D. Mass. 1995).

39 RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW § 207 (1987).40 See LORI F. DAMROSCH ET AL., INTERNATIONAL LAW: CASES AND MATERIALS

1037–45 (2009) (listing criminal sanctions against individuals, but not civil damages,as a human rights compliance mechanism under international law). There is a sensein which international law may be said to require that the victim be compensated. Ifconduct of an individual attributable to the state has caused injury to a citizen ofanother state, then the other state’s right to pursue a claim against the first state at



when the question is the existence of a cause of action for damages,the issue is not whether international law is to be accorded the forceof federal law, but, rather, whether international law is to be supple-mented with a cause of action for damages under domestic law.41 Aconclusion that a cause of action for damages does not exist underU.S. law for violation of a norm of customary international law doesnot establish that the norm lacks the status of federal law any morethan the lack of a private right of action to enforce a federal statute42

means that the federal statute lacks the force of federal law.Nor is it necessary to establish that customary international law has

the status of federal law in order to prevail in these cases. The SecondCircuit in Filartiga relied on the federal status of customary interna-tional law in finding § 1350’s grant of jurisdiction to be within thescope of Article III.43 Since Article III does not authorize diversityjurisdiction in suits between aliens, the court had to find that theaction “arose under” federal law. But § 1350 could be upheld underArticle III’s Arising Under Clause even if customary international lawwere not considered federal law. First, federal jurisdiction would beconsistent with Article III if the cause of action were created by federal

the international plane is sometimes subject to the precondition that the individualhave sought and been denied remedies through procedures available in the organs ofthe first state (including its courts). RESTATEMENT (THIRD) OF FOREIGN RELATIONS

LAW § 713 cmt. f. Thus, to avoid potential international responsibility, the offendingstate is required to provide the individual with a remedy. But the remedy need not bein the form of a damage judgment against the individual who inflicted the injury. Forexample, a direct payment by the state to the injured party might suffice. See id. § 901cmts. d, e; id. § 902 cmts. d–i. Thus, even if one understood international law torequire the offending state to provide to the victim the compensation that would avertits international responsibility to the state of the victim’s nationality, it could not besaid that international law of its own force entitled the victim to damages from the theindividual whose actions gave rise to the violation. In any event, this theory for findinga right of action for damages in international law would support a damage action onlyfor suits challenging conduct that would be attributable to the United States. Theconduct giving rise to the claim in Filartiga was not attributable to the United States.

41 But cf. Anthony J. Bellia Jr., & Bradford R. Clark, The Alien Tort Statute and theLaw of Nations, 78 U. CHI. L. REV. 445 (2011) (arguing that, at the time of the enact-ment of § 1350, the law of nations required nations to remedy injuries caused by theircitizens to citizens of other nations and that, therefore, a U.S. citizen’s tortious injuryto an alien would have been a “tort . . . in violation of the law of nations”). Comparethe similar argument discussed supra note 40 that international law under certain Rcircumstances indirectly obligates persons who have injured an alien to compensatethe alien.

42 See, e.g., Alexander v. Sandoval, 532 U.S. 275, 286–87 (2001).43 Filartiga, 630 F.2d at 885.



statute or treaty or federal common law.44 Second, jurisdiction couldsatisfy Article III even if the cause of action were not conferred byfederal law (for example, if it came from the law of the forum State orof the state in which the conduct took place). The Supreme Courtinterpreted Article III’s “arising under” provision very broadly inOsborn v. Bank of the United States,45 which has been read to permitgrants of jurisdiction over cases in which there is even a remote possi-bility that an issue of federal law might arise.46 The possibility that anissue of federal law will arise in a suit involving customary interna-tional law (even assuming that customary international law is not itselffederal)—for example, issues about the allocation of powers relatingto foreign relations47—is hardly remote.

It is true that the Court has distanced itself from this broad read-ing of Osborn,48 and scholars have challenged it.49 But § 1350 couldbe upheld under narrower interpretations of Article III’s ArisingUnder Clause even without recognizing a federal common law causeof action. Professor Wechsler argued that Congress may confer juris-diction under this clause over a category of cases if it has the power tolegislate substantively over that category, even without actually legislat-ing substantively.50 This test is easily satisfied by § 1350, since Article Igives Congress the power to legislate with respect to customary inter-

44 See American Well Works Co. v. Lane & Bowler Co., 241 U.S. 257, 260 (1916)(“A suit arises under the law that creates the cause of action.”). The Supreme Courteventually held that the cause of action in cases within the scope of § 1350 comesfrom federal common law. Sosa v. Alvarez-Machain, 542 U.S. 692, 732 (2004),

45 22 U.S. (9 Wheat.) 738, 759–63 (1824).46 See Verlinden B.V. v. Cent. Bank of Nigeria, 461 U.S. 480, 492–93 (1983); Tex-

tile Workers Union of Am. v. Lincoln Mills of Ala., 353 U.S. 448, 481–82 (1957)(Frankfurter, J., dissenting).

47 Bellia and Clark argue, and I agree, that the preemptive force of at least somenorms of customary international law derives from the constitutional allocation ofpowers regarding foreign relations. See infra Part II.D. My point here is that, even ifthis argument were rejected, the likelihood that a case involving customary interna-tional law will raise constitutional questions is enough to validate § 1350 under theOsborn interpretation of Article III.

48 See Mesa v. California, 489 U.S. 121, 136–37 (1989); Verlinden, 461 U.S. at492–93; see also Lincoln Mills, 353 U.S. at 481–82 (Frankfurter, J., dissenting).

49 See Anthony J. Bellia Jr., The Origins of Article III “Arising Under” Jurisdiction, 57DUKE L.J. 263, 334–37 (2007); Ernest A. Young, Stalking the Yeti: Protective Jurisdiction,Foreign Affairs Removal, and Complete Preemption, 95 CALIF. L. REV. 1775, 1820 (2007).

50 See Herbert Wechsler, Federal Jurisdiction and the Revision of the Judicial Code, 13LAW & CONTEMP. PROBS. 216, 224–25 (1948). For a defense of Wechsler’s reasoning,see Carlos M. Vazquez, The Federal “Claim” in the District Courts: Osborn, Verlinden,and Protective Jurisdiction, 95 CALIF. L. REV. 1731 (2007).



national law.51 Even more narrowly, Professor Mishkin has arguedthat Congress may grant jurisdiction over all cases arising in an area ofthe law in which the federal government has an “articulated and active. . . policy.”52 This standard too is easily satisfied by § 1350.53

Although scholars have rejected even these narrower theories of pro-tective jurisdiction,54 Supreme Court decisions upholding certaingrants of jurisdiction are hard to explain without resort to them,55 andthe Court itself has never rejected them.56 In short, even without afederal cause of action, the Article III question would turn on the sta-tus of customary international law as federal law (a) only with respectto a subset of § 1350 cases (those between aliens) and (b) only if theCourt were to reject a range of theories of Article III “arising under”jurisdiction proffered by jurists of the caliber of John Marshall, Her-bert Wechsler, and Paul Mishkin. The relevance of the status of cus-tomary international law to the Article III question in litigation under§ 1350 thus has always had a distinctly hypothetical cast.

In sum, because the status of customary international law as fed-eral law is neither necessary nor sufficient to resolve the key issues incases under § 1350, a focus on this statute is unhelpful. I will focusinstead on cases that present the status issue more directly. Perhapsthe best test case is a claim under State law against an individual whoclaims an immunity under international law. Some international lawimmunities are now addressed by statute57 or treaty,58 but the immu-

51 See U.S. CONST. art. I, § 8, cl. 10 (“The Congress shall have Power . . . [t]odefine and punish . . . Offences against the Law of Nations.”).

52 Paul J. Mishkin, The Federal “Question” in the District Courts, 53 COLUM. L. REV.157, 192 (1953).

53 See Carlos M. Vazquez, Comment, Verlinden B.V. v. Central Bank of Nigeria:Federal Jurisdiction over Cases Between Aliens and Foreign States, 82 COLUM. L. REV. 1057,1081 (1982).

54 See Young, supra note 49, at 1779–81. But see Vazquez, supra note 50 (defend- Ring protective jurisdiction).

55 See RICHARD H. FALLON, JR. ET AL., HART & WECHSLER’S THE FEDERAL COURTS

AND THE FEDERAL SYSTEM 765–68 (6th ed. 2009) (discussing bankruptcy cases) [here-inafter HART & WECHSLER].

56 Cf. Textile Workers Union of Am. v. Lincoln Mills of Ala., 353 U.S. 448, 460(1957) (Burton, J., concurring) (endorsing protective jurisdiction).

57 See Foreign Sovereign Immunities Act of 1976, Pub. L. No. 94-583, 90 Stat.2891 (codified in scattered sections of 28 U.S.C.).

58 See Vienna Convention on Diplomatic Relations arts. 29–40, Apr. 18, 1961, 23U.S.T. 3227, 500 U.N.T.S. 95; Vienna Convention on Consular Relations arts. 41, 43,53, Apr. 24, 1963, 21 U.S.T. 77, 596 U.N.T.S. 261.



nity of heads of state and foreign state officials is not.59 When animmunity applies, international law itself requires domestic courts torefrain from exercising jurisdiction. Customary international law doesall of the necessary work, and the only questions are whether interna-tional law has the status of domestic law, and, if so, whether it is fed-eral or State law.60

Second, for related reasons, I shall not address what revisionistsregard as one of the key tenets of the modern position: that customaryinternational law is self-executing federal law for purposes of the fed-eral courts’ “arising under” jurisdiction.61 The jurisdiction of thelower federal courts is itself always non-self-executing in the sense thatCongress must confer it by statute. As we have seen, under currentdoctrine, the Arising Under Clause permits Congress to confer suchjurisdiction over cases involving customary international law even ifsuch law is not federal.62 Whether customary international law hasthe status of federal law might determine whether a lower federalcourt possesses jurisdiction under the current general federal ques-tion statute, but that statute would deny jurisdiction over many casesimplicating customary international law even if such law were deemedfederal,63 and it would authorize jurisdiction in certain circumstanceseven if such law were not deemed federal.64 The more important

59 See Samantar v. Yousef, 130 S. Ct. 2278 (2010); Michael A. Tunks, Diplomats orDefendants? Defining the Future of Head-of-State Immunity, 52 DUKE L.J. 651, 666–73(2002).

60 Of course, as discussed below, the answers to these questions may be quitecomplex. See infra text accompanying notes 195–205, 538–557. R

61 See, e.g., Bradley & Goldsmith, Critique, supra note 9, at 858. R62 Bradley and Goldsmith recognize this possibility. See id. at 873 & n.354.63 Since international law rarely establishes a remedy except against the state

itself, which will often be immune, customary international law will usually enter thecase as a defense, or as a basis for a particular construction of a State or federal stat-ute. In such circumstances, federal jurisdiction cannot be based on the existence ofan issue of customary international law because of the well-pleaded complaint rule.See Louisville & Nashville R.R. Co. v. Mottley, 211 U.S. 149, 152–54 (1908).

64 For example, jurisdiction would exist if the cause of action were conferred byfederal statute, or by federal common law. Cf. Sosa v. Alvarez-Machain, 542 U.S. 692,732 (2004) (recognizing a federal common law cause of action for violation of somenorms of customary international law). The Court in Sosa denied that its holdingimplied that § 1331 could support the creation of federal common law in the sameway that § 1350 can. See id. at 731 n.19. This does not mean, however, that jurisdic-tion under § 1331 would be lacking over the federal common law cause of action thatthe Court recognized in Sosa. It is well established that § 1331 confers jurisdictionover federal common law causes of action. See Illinois v. City of Milwaukee, 406 U.S.91, 100 (1972) (“We . . . conclude that § 1331 jurisdiction will support claims foundedupon federal common law as well as those of a statutory origin.”); 13D CHARLES ALAN

WRIGHT ET AL., FEDERAL PRACTICE AND PROCEDURE §3563 (3d ed. 2008) (“[C]laims



question is whether the Supreme Court may review the State courts’interpretations of customary international law. But here, again, theanswer does not turn on Article III. The Supreme Court may be givenappellate jurisdiction over some cases involving customary interna-tional law under the diversity clauses of Article III,65 and perhaps overall such cases under a protective jurisdiction interpretation of the“arising under” clause, but that does not tell us whether the Courtmay review and reverse a State court’s application of such law. Thelatter question is governed by Article VI, not Article III. If customaryinternational law were purely a matter of State law, the federal courts(including the Supreme Court) would presumably have to treat theState courts’ interpretations of it as authoritative.66 My analysisaccordingly focuses on whether, and to what extent, federal law limitsState discretion to violate or depart from customary international law.

Third, unlike revisionists and defenders of the modern positionalike, I shall not refer to the modern position as the claim that cus-tomary international law has the status of federal common law, as Ithink that that label is unhelpful and potentially misleading. Instead,I shall describe the modern position as according to customary inter-national law (or some subset thereof) the status of preemptive federallaw. The canonical definition of federal common law is found in Hart& Wechsler’s The Federal Courts and the Federal System, which defines it as

based upon federal common law invoke federal-question jurisdiction under[§ 1331].”). For the same reason, I disagree with scholars who have read Justice Sou-ter’s opinion in Sosa as denying that customary international law is federal law forpurposes of § 1331. See Bellia & Clark, supra note 41, at 549; David J. Bederman, Law Rof the Land, Law of the Sea: The Lost Link Between Customary International Law and GeneralMaritime Law, 51 VA. J. INT’L L. 299, 345 (2010).

65 See Bradford R. Clark, The Supremacy Clause as a Constraint on Federal Power, 71GEO. WASH. L. REV. 91, 103–04 (2003); Clark, supra note 24, at 1333–34. R

66 Under current doctrine, the Supreme Court would be able to review a Statecourt’s application of State law to determine if it had “fair or substantial support,” butonly if the State court’s interpretation of such law was antecedent to the court’s denialof a federal right. See HART & WECHSLER, supra note 55, at 462–63. The general rule Rbarring Supreme Court review of State law reflects a constitutional recognition of theState courts as the authoritative interpreters of their own laws. See, e.g., Mullaney v.Wilbur, 421 U.S. 684, 691 (1975) (“This Court . . . repeatedly has held that [S]tatecourts are the ultimate expositors of [S]tate law, and that we are bound by their con-structions except in extreme circumstances . . . .” (citations omitted)). The limitedexception for antecedent State law grounds is based on the Supremacy Clause andensures that the State courts are not manipulating their laws to defeat federal rights.See HART & WECHSLER, supra note 55, at 462–63. But cf. Jonathan F. Mitchell, Recon- Rsidering Murdock: State-Law Reversals as Constitutional Avoidance, 77 U. CHI. L. REV.1335 (2010) (arguing that Article III permits broader Supreme Court review of Statecourt judgments based on State law).



“federal rules of decision whose content cannot be traced directly bytraditional methods of interpretation to federal statutory or constitu-tional commands.”67 If customary international law were accordedthe status of federal law, then it would qualify as federal common lawunder this definition because the content of its norms would not bedirectly traceable to the Constitution or federal statutes or treaties.But, as Hart and Wechsler’s successors recognize, federal common lawis not a unitary category.68 Norms might be accorded the force offederal common law for different reasons, and the effect of normsgiven that label may well differ. The concept of federal common lawis thus unhelpful in justifying the federal status of any given set ofnorms. Use of the “federal common law” label in examining the sta-tus of customary international law invites unreflective and inappropri-ate application to customary international law of doctrine developedfor quite different sorts of “federal common law.” For example, Brad-ley, Goldsmith, Young, and others criticize the modern position asinconsistent with the established notion that federal common law isproper only if applied in an interstitial way and hews closely to legisla-tively articulated policy.69 But, while the application of such limits inother areas may be warranted, these limits are inapposite to the struc-tural constitutional rationale defended here for according preemptiveforce to customary international law.

The term “federal common law” is also misleading in that it sug-gests a more creative role for the courts than the modern positioncontemplates. The term “federal common law” evokes the role thatState courts are now understood to play in elaborating the commonlaw of their respective States—a creative process in which the sub-stance of the law evolves based on the courts’ views of sound policy.Revisionists’ frequent description of the modern position as contem-plating “federal common lawmaking”70 suggests that courts applyingcustomary international law are similarly making the law up as they goalong. But that is not what the Restatement and its defenders envision.International law contains its own rules about how to identify custom-ary international law. Under these rules, the content of such law is aresult of the consistent practice of states performed out of a sense of

67 HART & WECHSLER, supra note 55, at 607. R

68 See id. (noting the topic of federal common law “has a miscellaneous quality”).69 See Bradley, Goldsmith & Moore, supra note 33, at 902, 926; Ramsey, supra R

note 17, at 558; Young, supra note 19, at 413–14. R

70 See Bradley & Goldsmith, Critique, supra note 9, at 857, 872 & n.352 (emphasis Radded); Young, supra note 19, at 413 (emphasis added). R



legal obligation (opinio juris).71 The role of U.S. courts applying cus-tomary international law is to ascertain those norms based on statepractice and opinio juris, not to elaborate the norms on the basis oftheir views of good policy.

Bradley and Goldsmith deride the claim that the courts would be“finding” or “discovering” the norms of customary international law,rather than “making” them, as hopelessly naıve. “This argument,”they say in rejecting a point by Professor Henkin similar to mine here,“assumes a sharp distinction between law-interpretation and lawmak-ing that cannot survive even the mildest of legal realist critiques.”72

But the point is not that customary international law is highly determi-nate; the point is that the courts will be resolving the indeterminaciesby reference to the actions and statements of others, not their ownviews of good policy. In fact, as Young reminds us, Erie itself calls forsuch a process to be followed by the federal courts in applying Statelaw.73 As we do not regard the federal courts’ role in applying State

71 For a recent study on this topic by the International Law Association, seeCOMM. ON FORMATION OF CUSTOMARY (GENERAL) INT’L LAW, INT’L LAW ASS’N, STATE-

MENT OF PRINCIPLES APPLICABLE TO THE FORMATION OF GENERAL CUSTOMARY INTERNA-

TIONAL LAW (2000) [hereinafter ILA STATEMENT OF PRINCIPLES], available at http://www.ila-hq.org/en/committees/index.cfm/cid/30 (follow “Conference ReportLondon 2000” hyperlink).

72 Bradley & Goldsmith, Critique, supra note 9, at 855. They continue: R

More importantly, it ignores the character of [customary international law]lawmaking: [Customary international law] is often unwritten, the necessaryscope and appropriate sources of “state practice” are unsettled, and therequirement that states follow customary norms from a “sense of legal obli-gation” is difficult to verify. Given what Professor Henkin himself refers to as[customary international law]’s “soft, indeterminate character,” it makes nosense to say that judges “discover” an objectively identifiable [customaryinternational law]. In fact, the process of identifying and applying [custom-ary international law] is at least as subjective as the domestic common lawprocess. This is particularly true of the new [customary international law],which is less tied than traditional [customary international law] to “objec-tive” evidence of state practice.

Id. (footnotes omitted) (quoting LOUIS HENKIN, INTERNATIONAL LAW: POLITICS AND

VALUES 29 (1995)).73 Additionally, as discussed in Part III, if the indeterminacy of customary interna-

tional law were a concern, the problem could be addressed by according the force ofpreemptive federal law only to norms of customary international law that satisfy ahigher standard of clarity and breadth of acceptance. The Court has taken thisapproach in two discrete areas, see infra notes 184–204, 237–242 and accompanying Rtext, and it could defensibly extend it to other contexts, see infra notes 558-587 and Raccompanying text.



law as “lawmaking,” neither should we view their role in applying cus-tomary international law that way.74

Even if the norms themselves are not “made” by the courts thatapply them, one might say that the decision to accord customary inter-national law the force of preemptive federal law is itself an act of judi-cial lawmaking. But, as explained by Bellia and Clark, and elaboratedherein, the preemptive force of (at least some) customary interna-tional law can be understood as a valid implication from the constitu-tional structure.75 Revisionists might reject the inference on themerits, but presumably they do not deny that inference from structureis a valid approach to constitutional interpretation.76 If the decisionto accord norms of international law the status of federal law weremade at retail, as Young proposes,77 the resulting regime would in factresemble one of judicial lawmaking. In theory, a retail approachcould turn on the factors made relevant by the structural approachdefended here, such as the danger of interference with the politicalbranches’ conduct of foreign relations likely to result from the States’violation of particular norms. But, because courts lack the relevantnorm-specific information and expertise, a retail approach wouldlikely degenerate into ad hoc decisionmaking reflecting the particularjudges’ views of the merits of particular norms.78 Concerns about

74 This is not to deny that indeterminacy of norms leaves judges with wider inter-pretive discretion. Cf. infra Part III (discussing possible responses to increased inde-terminacy of customary international law).

75 Some defenders of the modern position advance originalist arguments regard-ing the meaning of the term “laws of the United States” in Article III and/or ArticleVI. See, e.g., Jordan J. Paust, International Law Before the Supreme Court: A Mixed Record ofRecognition, 45 SANTA CLARA L. REV. 829, 832–33 (2005) (“[V]arious Founders,Supreme Court Justices, and other federal judges have long recognized more gener-ally that . . . the phrase ‘laws of the United States,’ encompasses the law of nations orwhat we term ‘customary international law.’”). I do not pursue those arguments here.Although I argue below that the constitutional text does not establish the revisionistposition, see infra text accompanying notes 324–342, I also do not think that the text Ralone clinches the modern position, given the possibility that some Founders mayhave understood the law of nations to have the status of general common law. I do,however, advance an originalist argument regarding the constitutional structure,which I think provides stronger support for the modern position. See infra textaccompanying notes 117, 341–352 (discussing Ware v. Hylton, 3 U.S. (3 Dall.) 199 R(1796) and Rutgers v. Waddington). And structural arguments are necessary to supportthe argument that, to the extent that customary international law was understood aspre-Erie general common law, it should now be regarded as preemptive federal law.

76 See CHARLES L. BLACK, JR., STRUCTURE AND RELATIONSHIP IN CONSTITUTIONAL

LAW 7–32 (1969).77 See Young, supra note 34, at 29–30. R78 Note that we are assuming here that the norm exists under international law

and that the court is deciding whether to accord it the status of U.S. law.



judicial lawmaking thus cut in favor of an approach that reads theConstitution as according federal status to customary international lawas a whole, or to some substantial subset determined by generallyapplicable standards, such as the degree of clarity or the breadth of itsacceptance as a norm of customary international law. That, in anyevent, is the version of the modern position that I shall defend here.Under such a scheme, the norms themselves are not made by thecourts,79 and the determination whether the norms have the force offederal law is no more an act of judicial lawmaking than is the resolu-tion of any other constitutional question on structural grounds.80

Finally, even though the status of customary international law is aconstitutional question in important respects, it is proper to place aheavy burden of persuasion on those seeking to depart from the “well-entrenched”81 modern position. It is true that the Supreme Courtapplies a weaker version of stare decisis to constitutional questions.82

But the reason for the Court’s greater (though still limited83) willing-ness to reject settled law on constitutional questions is that the courts’errors cannot be corrected by Congress.84 For stare decisis purposes,the status of customary international law should be treated as a sub-

79 As discussed in Part IV, decisions of domestic courts interpreting customaryinternational law do, over time and in conjunction with the acts of many others, con-tribute to the formation of new rules or the dissolution of old rules of customaryinternational law. But this phenomenon hardly rises to the level of “lawmaking” bythose courts. In any event, it is not the sort of lawmaking that concerns therevisionists.

80 For related reasons, I quibble with the locution of some revisionists anddefenders of the modern position who say that, under the modern position, the courtsincorporate customary international law as U.S. law, either at retail or wholesale. See,e.g., Koh, supra note 16, at 1827. As I defend the modern position here, the courts do Rnot incorporate such norms into U.S. law; rather, they read the Constitution asaccording such norms (or some subset thereof) the force of preemptive federal law.Again, I do not suggest that the act of constitutional interpretation is mechanical, butI do think it is useful to distinguish the courts’ exercise of judgment from their exer-cise of will.

81 Bradley & Goldsmith, Critique, supra note 9, at 816. R82 See Payne v. Tennessee, 501 U.S. 808, 828 (1991); Christopher P. Banks, The

Supreme Court and Precedent: An Analysis of Natural Courts and Reversal Trends, 75 JUDICA-

TURE 262, 263 tbl.1 (1992) (showing that 60.5 percent of overturned precedent from1789 to 1991 was constitutional and 27 percent statutory); David S. Cohen, The Prece-dent-Based Voting Paradox, 90 B.U. L. REV. 183, 220–21 (2010).

83 See, e.g., Planned Parenthood of Se. Pa. v. Casey, 505 U.S. 833, 844, 860–61(1992).

84 See Payne, 501 U.S. at 828 (“Stare decisis is not an inexorable command . . .particularly . . . in constitutional cases, because in such cases ‘correction throughlegislative action is practically impossible.’” (quoting Burnet v. Coronado Oil & GasCo., 285 U.S. 393, 407 (1932) (Brandeis, J., dissenting))).



constitutional issue. Both the modern and the revisionist positionsare merely default rules. With limited exceptions, whichever rulewere adopted would be subject to revision, either at retail or whole-sale, by statute or treaty or even sometimes by the Executive alone.85

Since the issue is within the control of the federal political branches,the stronger version of stare decisis that is applied to statutory issues isthe pertinent one.86

85 One possible exception would be where the federal government’s power to actturns on customary international law, as may be the case when Congress legislatesunder the Offenses Clause. Compare Michael Stokes Paulsen, The Constitutional Powerto Interpret International Law, 118 YALE L.J. 1762, 1808 (2009) (arguing that theOffenses Clause empowers Congress to enact its own interpretation of internationallaw as a matter of federal law), with Stephen I. Vladeck, The Laws of War as a Constitu-tional Limit on Military Jurisdiction, 4 J. NAT’L SECURITY L. & POL’Y 295, 299 & n.27(2010) (arguing that customary international law functions as a limit on Congress’spower to legislate pursuant to the Offenses Clause). A possible exception to Con-gress’s power to repeal customary international law would be where such repealwould constitute an unconstitutional interference with the President’s exclusive pow-ers. See Lewis Yelin, Head of State Immunity as Sole Executive Lawmaking, 45 VAND. J.TRANSNAT’L L. (forthcoming 2011) (on file with author) (arguing on this ground thatCongress would lack the power to repeal the customary international law of head-of-state immunity).

86 I have made this point with respect to the related doctrine relating to the dor-mant foreign affairs power. See Carlos Manuel Vazquez, W(h)ither Zschernig?, 46 VILL.L. REV. 1259, 1308 (2001) [hereinafter Vazquez, W(h)ither Zschernig?]. Because Con-gress retains control, the question discussed herein might be regarded as one of con-stitutional common law, rather than of constitutional law proper. See Henry P.Monaghan, Foreword: Constitutional Common Law, 89 HARV. L. REV. 1, 2–3 (1975); seealso Alfred Hill, The Law-Making Power of the Federal Courts: Constitutional Preemption, 67COLUM. L. REV. 1024, 1025 (1967).

A brief word on the Court’s recent decision in Medellın v. Texas, 552 U.S. 491(2008), is also warranted. The Court in Medellın found that the treaty before it wasnot self-executing and accordingly did not preempt an inconsistent State law. Exactlywhy the Court found that the treaty was not self-executing has been the subject ofmuch debate. According to some commentators, the Court found that the treatyimposed a firm obligation on the United States but was non-self-executing because itdid not clearly and affirmatively state that it was intended to have the force of domes-tic law. See, e.g., Julian G. Ku, Medellın’s Clear Statement Rule: A Solution for Interna-tional Delegations, 77 FORDHAM L. REV. 609, 609 (2008). If this is what the Court held,then Medellın may be thought to have rejected the modern position by implication. Ifa treaty that imposes a firm international obligation does not preempt inconsistentState law unless it also affirmatively provides that it has the force of domestic law, eventhough the Constitution declares treaties to be “the supreme Law of the Land,” U.S.CONST., art. VI, cl. 2, then it may follow that customary international law does notpreempt State law in the absence of an act of domestic lawmaking giving it preemp-tive effect. This reading of Medellın, however, is highly contested. Even scholars whoagree with the Court’s holding do not read Medellın to have held that treaties that aresilent as to their domestic effect are for that reason alone non-self-executing. See Cur-



I. THE MODERN POSITION: EXPLICATION AND PRELIMINARY DEFENSE

Critics of the modern position describe it as the claim that “cus-tomary international law is always federal law,”87 or that “all of [cus-tomary international law], ‘whatever [it] requires,’ is automaticallyincorporated wholesale into post-Erie federal common law.”88 As dis-cussed more fully below, the modern position has never been souncompromising.89 Specifically, the modern position has never main-tained that all of customary international law is federal law insofar as itis sought to be applied to invalidate the acts of foreign states or of thefederal government. Adherents of the modern position take varyingpositions regarding the circumstances in which customary interna-tional law is federal law binding on such actors. What unites them isthe claim that customary international law is binding upon the Statesand that State laws and conduct that would violate customary interna-tional law are thus invalid. The version of the modern position that Idefend here is that customary international law preempts State lawinsofar as it imposes obligations on State officials and private parties.Insofar as it is preemptive federal law, moreover, the lower federalcourts may be given jurisdiction over cases “arising under” it, and the

tis A. Bradley, Self-Execution and Treaty Duality, 2008 SUP. CT. REV. 131, 178–79; DavidH. Moore, Law(makers) of the Land: The Doctrine of Treaty Non-Self-Execution, 122 HARV.L. REV. F. 32, 46–47 (2009), http://www.harvardlawreview.org/media/pdf/moore.pdf; Ernest A. Young, Treaties as “Part of Our Law,” 88 TEX. L. REV. 91, 118, 123 (2009;see also John T. Parry, Response, Rewriting the Roberts Court’s Law of Treaties, 88 TEX. L.REV. SEE ALSO 65 (2010), http://www.texaslrev.com/sites/default/files/seealso/vol88/pdf/88TexasLRevSeeAlso65.pdf (noting that very few defenders of Medellınadopt the broad reading noted above). I have offered two alternative readings ofMedellın. See Carlos M. Vazquez, Treaties as Law of the Land: The Supremacy Clause andthe Judicial Enforcement of Treaties, 122 HARV. L. REV. 599, 662–63 (2008) (arguing thatMedellın is best read to have interpreted the relevant treaty as not imposing a firmobligation under international law) [hereinafter Vazquez, Treaties as Law of the Land];Carlos Manuel Vazquez, Less than Zero?, 102 AM. J. INT’L L. 563, 572 (2008) [hereinaf-ter Vazquez, Less than Zero?] (arguing that Medellın could be read to hold that treatiesare non-self-executing only if they affirmatively provide that they are subject to legisla-tive implementation). Either reading is consistent with the modern position. Ratherthan read Medellın as ruling out the modern position, I regard the authorities cited inthis Article for treating customary international law as preemptive of State law as addi-tional reasons for rejecting the broad reading of Medellın described above.

87 Young, supra note 19, at 437. R

88 Bradley, Goldsmith & Moore, supra note 33, at 902 (second alteration in origi- Rnal) (quoting Lea Brilmayer, Federalism, State Authority, and the Preemptive Power of Inter-national Law, 1994 SUP. CT. REV. 295, 324).

89 See infra text accompanying notes 206–233. R



Supreme Court may be given jurisdiction to review State court applica-tions and interpretations of it.90

A. The Basic Case for the Modern Position

The canonical expression of the modern position is the statementin The Paquete Habana91 that “[i]nternational law is part of our law,and must be ascertained and administered by the courts of justice ofappropriate jurisdiction, as often as questions of right dependingupon it are duly presented for their determination.”92 Revisionists dis-miss this and many similar statements by the Supreme Court in itsmany cases applying customary international law93 as merely reflectinga pre-Erie consensus that customary international law was part of thegeneral common law. Since Erie overthrew the concept of generalcommon law, this status is no longer available. Like the rest of thecommon law, they contend, customary international law should nowbe regarded as, at best, State law.

This subpart sets forth the basic case for treating customary inter-national law, post-Erie, as preemptive of State law. I respond here tosome of the arguments advanced by the first wave of revisionists, but Idefer to Part II my responses to the arguments of the proponents ofthe intermediate positions. In this Part, I defend the broad view thatall of customary international law is binding on State actors as thebetter modern translation of the constitutional structure as under-stood by the Founders, as well as of pre-Erie doctrine. In Part III, Iconsider whether recent developments in the nature or content ofcustomary international law justify a shift in our approach to such law.

90 The Supreme Court’s jurisdiction is conferred directly by Article III of theConstitution, but Congress has the power to make “exceptions and regulations” withrespect to such jurisdiction, and Section 25 of the Judiciary Act of 1789 has beeninterpreted to exclude from the Supreme Court’s appellate jurisdiction any cases notauthorized by that section. Ex Parte McCardle, 74 U.S. (7 Wall.) 506, 512–13 (1868).“The principle that the affirmation of appellate jurisdiction implies the negation ofall such jurisdiction not affirmed having been thus established, it was an almost neces-sary consequence that acts of Congress, providing for the exercise of jurisdiction,should come to be spoken of as acts granting jurisdiction, and not as acts makingexceptions to the constitutional grant of it.” Id. at 513.

91 175 U.S. 677 (1900).92 Id. at 700.93 See generally David J. Bederman, Customary International Law in the Supreme Court,

1861–1900, in INTERNATIONAL LAW IN THE U.S. SUPREME COURT, supra note 34, at 89; RWilliam S. Dodge, Customary International Law in the Supreme Court, 1946–2000, inINTERNATIONAL LAW IN THE U.S. SUPREME COURT, supra note 34, at 353; Ramsey, supra Rnote 34. R



1. Constitutional Structure and Original Intent

The structural reason for treating customary international law asfederal law is straightforward. Violations of international law by theStates are attributable to the nation as a whole.94 Such violations thusjustify nations harmed by them to take retaliatory action against theUnited States as a whole. At the time of the Founding, retaliationcould take the form of military action.95 Today, military action inresponse to violations of international law is itself a violation of inter-national law, except for defensive action in response to anothernation’s unlawful use of force or military action taken pursuant toauthorization by the U.N. Security Council or, perhaps, in other lim-ited circumstances.96 Still, nations may respond to another nation’sviolation of international law by taking countermeasures not involvingthe use of force.97 Countermeasures can and usually will be directedat the nation as a whole, or States other than the one that committedthe violation.98 In addition to risking countermeasures against thenation, a State’s violation of international law can be expected to pro-duce international friction and an unfriendly attitude toward theUnited States on the part of injured or otherwise offended nations,which in turn can be expected to complicate the federal government’sefforts to achieve the nation’s foreign relations goals.

While the costs of a State’s violations of customary internationallaw will be borne by the nation as a whole, the benefits will usually beenjoyed by that State alone. Were States free to violate internationallaw, therefore, one would expect excessive violations, as the Stateswould be able to externalize the costs of the violations to their sisterStates. To put it differently, the rest of the nation would be subsi-dizing individual States’ decisions to violate international law. Thestructural reason for denying States the power to violate customaryinternational law is well captured in Hamilton’s statement that “the

94 See Draft Articles on Responsibility of States for Internationally Wrongful Actsart. 4(1) & cmt. 9, 56 U.N. GAOR Supp. No. 10, at 40, U.N. Doc. A/56/10 (2001)[hereinafter Responsibility of States], reprinted in [2001] 2 Y.B. Int’l L. Comm’n 26,U.N. Doc. A/CN.4/SER. A/2001/Add.1 (Part 2).

95 See Bellia & Clark, supra note 21, at 6. R

96 See U.N. Charter art. 2, para. 4.97 Responsibility of States, supra note 94, art. 22 & cmt.1. R

98 Cf. Peter J. Spiro, Foreign Relations Federalism, 70 U. COLO. L. REV. 1223,1261–62 (1999) (arguing that foreign states now sometimes retaliate against the spe-cific State that committed the violation).



peace of the WHOLE ought not to be left at the disposal of aPART.”99

A recent episode concerning the execution of Angel Breard illus-trates the structural problem.100 Breard was a Paraguayan nationalwho was convicted of attempted rape and murder in Virginia and sen-tenced to death.101 Paraguay initiated proceedings against the UnitedStates before the International Court of Justice (ICJ), claiming thatBreard’s execution would violate the United States’ obligations underthe Vienna Convention on Consular Relations.102 The ICJ, in turn,ordered that Breard’s execution be stayed pending its resolution ofParaguay’s claims.103 While Breard and Paraguay pursued relief in theU.S. courts, Secretary of State Madeleine Albright wrote a letter toVirginia Governor James Gilmore urging him to stay the execution.104

Albright told Gilmore that compliance with the ICJ’s order was impor-tant to the nation’s foreign relations and to the safety of Americancitizens abroad whose rights depended on the Vienna Convention.105

In response, the Governor issued a statement denying the Secretary ofState’s request because, “[a]s Governor of Virginia my first duty is toensure that those who reside within [Virginia’s] borders . . . may con-duct their lives free from the fear of crime.”106 The U.S. courts even-tually denied relief on dubious grounds107 and Breard was executedin defiance of the ICJ’s order.108 Governor Gilmore behaved perfectlyrationally. He was elected by the citizens of Virginia, not the citizensof New York or California or the forty-seven other States whoseresidents may have been traveling abroad and whose interests the Sec-retary of State believed would be harmed by Gilmore’s defiance of the

99 THE FEDERALIST NO. 80, at 476 (Alexander Hamilton) (Clinton Rossiter ed.,1961).100 This illustration involves obligations under treaties rather than customary

international law, but the structural problem is the same in the two contexts.101 Jonathan I. Charney & W. Michael Reisman, Agora: Breard, 92 AM. J. INT’L L.

666, 666 (1998).102 Id. at 669.103 Id. at 671.104 Id. at 671–72.105 Id. at 684.106 Id. at 674 (quoting Gov. Gilmore).107 Breard v. Greene, 523 U.S. 371 (1998) (per curiam); see Carlos Manuel Vaz-

quez, Night and Day: Coeur d’Alene, Breard, and the Unraveling of the Prospective-Retro-spective Distinction in Eleventh Amendment Doctrine, 87 GEO. L.J. 1, 51 (1998).108 Id. at 6.



ICJ. The structural problem with leaving the interests of the rest ofthe nation in Gilmore’s hands requires no elaboration.109

The Founders recognized this structural problem. The inabilityof the federal government to prevent or remedy State violations of thelaw of nations during the period of the Articles of Confederation wascited at the Convention as an important reason for adopting a newconstitution.110 Such violations during the period of the Articles ofConfederation troubled the Founders because they complicated thenational government’s ability to achieve its foreign policy goals and,indeed, threatened to plunge the nation into destructive wars.111 TheFounders were virtually unanimous in the view that the power overinternational law should be given to the federal government.

Revisionists do not deny this, but they argue that the constitu-tional text shows that the Founders’ solution to the problem, insofaras customary international law was concerned, was to give the powerto require compliance with such law to the lawmakers, who wereempowered to “define and punish offenses against the law ofnations,”112 and the treaty-makers, who were given the power to codifynorms of the law of nations in self-executing treaties.113 The Foun-ders’ failure to include the law of nations in the Supremacy Clause,however, shows that they did not regard such law as self-executing fed-eral law.114 Thus, the revisionists claim, the States remain free to vio-late such law in the absence of laws or treaties providing otherwise.

I consider the constitutional text’s conformity with the modernposition in Part II, as Ramsey advances the textual arguments mostforcefully, and I conclude that the text neither establishes nor rebuts

109 On why the Supreme Court’s decision in the Breard case deserves little, if any,weight as precedent, see id. at 66–68.110 See 1 THE RECORDS OF THE FEDERAL CONVENTION OF 1787, at 19 (Max Farrand

ed., rev. ed. 1937) (“[Edmund Randolph] then proceeded to enumerate the defects:. . . that particular [S]tates might by their conduct provoke war without controul. . . .”); id. at 24–25 (“[Randolph] observed that . . . [i]f a state acts against a foreignpower contrary to the laws of nations or violates a treaty, [the Confederation] cannotpunish that State, or compel its obedience to the treaty. . . . It therefore cannot pre-vent a war.”); JAMES MADISON, VICES OF THE POLITICAL SYSTEM OF THE UNITED STATES,(1787) reprinted in SELECTED WRITINGS OF JAMES MADISON 35, 35–36 (Ralph Ketchamed., 2006).111 See Stephens, supra note 16, at 402–03, 403 n.44. R

112 Bradley & Goldsmith, Critique, supra note 9, at 856; see RAMSEY, supra note 17, Rat 348–55.113 See RAMSEY, supra note 17, at 347–48; Bradley & Goldsmith, Critique, supra note R

9, at 819–20. R

114 See RAMSEY, supra note 17, at 348–55; Bradley & Goldsmith, Current Illegitimacy, Rsupra note 9, at 321. R



the modern position.115 Here, I merely note that treating the law ofnations as non-self-executing presents its own structural problem, aproblem emanating from another feature of our constitutional systemthat the revisionists themselves often invoke in objecting to federalcommon law. In order to protect the States from having their lawpreempted too casually, the Constitution imposes daunting require-ments for the enactment of federal law.116 Lawmaking requires theagreement of majorities of two differently constituted houses of thelegislature plus the agreement of a third independently elected offi-cial (the President), or, if the President does not agree, a two-thirdsvote of the two legislative bodies. Treaty-making requires the agree-ment of the President and two-thirds of one of the houses of the legis-lature. These requirements were imposed to place a heavy weight onthe scales in favor of the legal status quo. Given the Founders’ recog-nition of the clear structural problem that would result if States werefree to violate international law, it would be surprising if theyregarded the successful completion of the legislative or treaty-makinggauntlet as a predicate for federal enforcement of the law of nationsagainst the States.

And, in fact, the evidence shows that the Founders did not regardsuch enforcement to be subject to affirmative incorporation of cus-tomary international law in a statute or treaty. Some Founders under-stood that national control over compliance with international law wasa necessary implication of the constitutional structure. Indeed, evenbefore the adoption of the Constitution, State courts expressed theview that, in any federal system, the States must lack the power to vio-late international law. For example, in Rutgers v. Waddington, decidedduring the period of the Articles of Confederation, the Mayor’s Courtof New York wrote that it would be “contrary to the very nature of the[U.S.] confederacy” for a single state like New York “to abrogate oralter one of the known laws or usages of nations.”117

115 See infra notes 323–40 and accompanying text. R116 For the propositions in this paragraph, see Clark, supra note 24, and discussion R

infra note 474 and accompanying text. Clark argues that the procedural obstacles to Rfederal lawmaking were designed to protect the States from having their law pre-empted. Once State law has been preempted by federal law, however, the dauntingrequirements for enacting federal legislation actually make it difficult to devolvepower back to the States. Thus, it is more accurate to say that the daunting require-ments for enacting federal legislation operate to protect the status quo. See CarlosManuel Vazquez, The Separation of Powers as a Safeguard of Nationalism, 83 NOTRE DAME

L. REV. 1601 (2008). In the beginning, however, the status quo was that State law (orgeneral common law) governed most legal matters.117 SELECT CASES OF THE MAYOR’S COURT OF NEW YORK CITY, 1674–1784, at 316

(Richard B. Morris ed., 1935).



It is true that some statements of the Founders are consistent witha belief that the law of nations was part of the common law.118 Butthese statements do not support the claim that the Founders viewedcustomary international law as anything resembling modern-day Statelaw. First, at the time of the Founding, the law of nations was thoughtto include not just the law addressing the mutual rights and obliga-tions of states, but also the law merchant and general commerciallaw.119 Only the state-to-state branch of the law of nations includedrules whose violation justified resort to war or other countermea-sures.120 Thus, only this subset of the law of nations implicated thestructural rationale for treating customary international law as pre-emptive of State law. The other branches of the law of nations, bycontrast, were understood to be subject to local variation.121 Theycorrespond to modern-day conflict of laws or simply domestic law.Adherents of the modern position do not claim that the modern-daycounterparts of the latter branches of the law of nations have theforce of preemptive federal law. In referring to the law of nationsgenerally as part of the common law, the Founders likely did notmean to deny that a subset of it was preemptive of State law for struc-tural constitutional reasons.

Second, as recounted by Bellia and Clark, in the early days of theRepublic, it was widely believed that the common law had been incor-porated wholesale as national law.122 Although this position was ulti-mately rejected,123 the fact that it was widely held shows that at leastsome Founders who regarded the law of nations as common lawviewed it as federal law.

Third, even if some Founders regarded all of the law of nations ashaving common law status and did not regard the common law asfederal law, it does not follow that they understood the status of the

118 Attorney General Randolph stated, when deciding whether the arrest of a pub-lic minister’s servant could be punished under common law, that the law of nationswas incorporated into the “law of the land.” 1 Op. Att’y Gen. 27 (1792). In Henfield’sCase, 11 F. Cas. 1099 (C.C.D. Pa. 1793) (No. 6360), Justice James Wilson wrote that“though there has been no exercise of the power conferred upon congress by theconstitution ‘to define and punish offences against the laws of nations,’ the federaljudiciary has jurisdiction of an offence against the laws of nations, and may proceed topunish the offender according to the forms of the common law.” Id. at 1120 n.6. Forother statements by the Founders to this effect, see Stewart Jay, The Status of the Law ofNations in Early American Law, 42 VAND. L. REV. 819, 825–27 (1989).119 See Bellia & Clark, supra note 21, at 19–20. R120 See id. at 11.121 See id. at 20–22.122 See id. at 46–48.123 See id. at 55–56.



law of nations as anything like modern-day State law. The characteris-tics of the general law of the early Republic were very different fromthose of today’s State law. “[I]n early nineteenth century usage, ‘com-mon law’ was a general common law shared by the American statesrather than a local common law of a particular state.”124 Before Erie,the federal courts “exercised their independent judgment on whatthat law required.”125 Thus, by giving the federal courts constitutionaland statutory jurisdiction over the types of cases most likely to raiseissues under the law of nations,126 even Founders who believed thatthe law of nations merely had the status of the common law were ableto give effect to their “clear[ ] inten[t] that customary internationallaw should have application in our courts,”127 as well as their “well-documented desire to ensure that [S]tates complied with interna-tional law.”128 By contrast, federal courts today must follow the Statecourts’ decisions regarding what State law requires.

2. Pre-Erie Doctrine

Revisionists contend that, before Erie, customary internationallaw was understood to have the status of Swift-era general law. Assuch, States were understood to have the power to depart from suchlaw, and thus to place the United States in breach of it, and theSupreme Court lacked the power to review State court interpretationsand applications of it.129 But the cases cited by the revisionists do notestablish that States were thought to be free to violate internationallaw, and they are at best equivocal regarding the availability ofSupreme Court review. Moreover, as Bellia and Clark have shown,there were also cases during this era that articulated an allocation-of-powers rationale for treating a subset of the law of nations as preemp-tive federal law. As discussed in subpart II.D, this subset of the law ofnations appears to have comprised the state-to-state branch of the lawof nations, which is what we regard as customary international lawtoday. Even some of the cases cited by revisionists as denying theSupreme Court’s jurisdiction to review State court applications of cus-tomary international law recognize that such law was neverthelessbinding on the States. Revisionists make much of the absence of cases

124 William A. Fletcher, The General Common Law and Section 34 of the Judiciary Act of1789: The Example of Marine Insurance, 97 HARV. L. REV. 1513, 1515 n.9 (1984).125 Id. at 1515.126 Bellia & Clark, supra note 21, at 6. R127 Young, supra note 19, at 390. R128 Bradley, Goldsmith & Moore, supra note 33, at 883. R129 See Bradley & Goldsmith, Critique, supra note 9, at 824. R



holding State laws invalid on the ground that they violate customaryinternational law,130 but it is equally noteworthy that there appear tobe no cases upholding the States’ ability to depart from such law, andmuch dicta denying their power to do so.

Furthermore, even if the pre-Erie cases did regard customaryinternational law as Swift-era general common law, the general com-mon law was very different from today’s State law. The pre-Erieregime of general common law, in operation, resembled present-dayfederal law more than present-day State law. To put the matter mostfavorably to the revisionist position, pre-Erie doctrine reflected twocontending positions regarding the state-to-state branch of the law ofnations—some believed it to be preemptive federal law and othersbelieved it to be part of the general law. If the latter category is nolonger available, then pre-Erie doctrine supports a post-Erie status ofcustomary international law as federal law, not a State law status thathad never been in play.

a. State Authority to Depart from Customary InternationalLaw

Bradley and Goldsmith cite several early cases involving theimmunities of foreign officials as establishing that it was then under-stood that “one consequence of [customary international law]’s statusas general common law was that a [S]tate of the Union had the ability,in the absence of a constitutional provision or federal enactment tothe contrary, to violate international law and thereby implicate theinternational responsibility of the United States.”131 But the cases donot establish that the States were understood to have the power todepart from customary international law.

The “most notorious . . . example”132 that they discuss was NewYork’s prosecution of a British national, Alexander McLeod, in con-nection with the destruction of the American steamer, The Caroline.The British government claimed that McLeod was entitled to immu-nity from prosecution under international law.133 Bradley and Gold-smith assert that “Secretary of State Daniel Webster disclaimed theconstitutional power to release McLeod but accepted responsibilityfor any violation of the law of nations.”134 But, in fact, Webster only

130 See RAMSEY, supra note 17, at 352; Young, supra note 19, at 383. R131 Bradley & Goldsmith, Critique, supra note 9, at 825. R132 Id. at 825 n.56.133 Id.134 Id. (citing David J. Bederman, The Cautionary Tale of Alexander McLeod: Superior

Orders and the American Writ of Habeas Corpus, 41 EMORY L.J. 515, 515–20 (1992)).



denied that the Executive had the power to release McLeod. Webstermaintained that the State courts would apply the law of nations andthat, if they misapplied it, review would be available in the SupremeCourt.135 When McLeod sought habeas relief in the New York courts,the court did apply the law of nations, but found McLeod not to beentitled to immunity and denied him habeas relief.136 McLeod wasthen advised to appeal to the U.S. Supreme Court, but he declined todo so because of the delay involved.137 Instead, he submitted to a jurytrial and was acquitted.

The other cases cited by Bradley and Goldsmith for the proposi-tion that States were understood to have the power to depart fromcustomary international law are similarly consistent with the obliga-tion of State courts to apply such law and the power of the SupremeCourt to review State court applications of such law.138 These exam-

135 See Bederman, supra note 134, at 520 (citing Letter from Mr. Webster to Mr. RCrittenden (Mar. 15, 1841), reprinted in 29 BRITISH AND FOREIGN STATE PAPERS

1840–1841, at 1139, 1141–42 (London, Ridgway & Sons 1857)); see also Caperton v.Bowyer, 81 U.S. (14 Wall.) 216, 225–26 (1871) (argument of counsel for plaintiff inerror) (“That the understanding of the administration and of well-informed personsat the time [regarding the McLeod case] was, that this, the Supreme Court of theUnited States, had jurisdiction of the judgment of the Supreme Court of New York, isclear from a notice of the trial of McLeod in the National Intelligencer of May 22d,1841. The editors alluding to this trial there say: ‘Whatever the decision, whether forreleasing or remanding the prisoner, an appeal will probably be taken to the Court ofErrors, from which a further appeal lies, in cases of this nature, to the Supreme Courtof the United States.’ Mr. Choate, in a speech delivered on the 11th of June, 1841, inthe United States Senate, maintains the same position. He says: ‘The clear course ofthe government, therefore, was to do what it did, to have McLeod’s case fairly tried,and if needful, to have his case brought into the National tribunals.’” (footnote omit-ted) (citing NAT’L INTELLIGENCER, June 17, 1841)); cf. infra note 164 (addressing the Rresponse of counsel for defendant in error in Caperton). The Supreme Court inCaperton did not reach the issue.136 See Bederman, supra note 134, at 521–26. R137 See id. at 526; R.Y. Jennings, The Caroline and McLeod Cases, 32 AM. J. INT’L L.

82, 95 (1938).138 Bradley and Goldsmith cite the cases of General Collot and Don Joseph de

Cabrera, as recounted by Bederman, supra note 134, at 526–27. See Bradley & Gold- Rsmith, Critique, supra note 9, at 825 n.56. The case of General Collot establishes only Rthat the Executive cannot set a State prisoner free. See Bederman, supra note 134, at R526. Like McLeod, Collot was able to raise his defense under the law of nations in theState courts and prevailed without the need for Supreme Court review. Id. DonJoseph de Cabrera was denied federal habeas relief on the ground that no statuteauthorized such relief for persons held in the custody of the States, id. at 527, a hold-ing with no implications for the modern position, see Ex parte Bollman, 8 U.S. (4Cranch) 75, 93–94, 97–98 (1807). Nor can one draw any relevant inferences fromcases in which States “failed to prosecute the perpetrators of mob violence” eventhough prosecution was allegedly required by customary international law. Bradley &



ples do show that the federal government was not given unlimitedpower to force States to comply with customary international law. Butadherents of the modern position do not contend that the federalExecutive may unilaterally release a State prisoner in order to complywith customary international law,139 or that a lower federal court maygrant habeas relief to State prisoners without statutory authoriza-tion.140 Adherents of the modern position maintain, and revisionistsdeny, that customary international law is binding on State actors andreviewable in the Supreme Court. The cases cited by Bradley andGoldsmith are, if anything, more supportive of the modern positionthan of the revisionist position.

b. Supreme Court Review of Customary International Law

Revisionists also maintain that, before Erie, the law was clear thatthe Supreme Court lacked jurisdiction to review State court decisionsregarding the law of nations because such law was regarded asnonfederal law. But, as suggested by the McLeod incident just dis-cussed, the law on this point does not support the revisionists asclearly as they contend. In fact, the cases are at best equivocal and,indeed, contain support for the view that customary international lawis binding on the States. The cases as a whole show, at most, thatthere were two contending positions during this era: that the state-to-

Goldsmith, Critique, supra note 9, at 825. Indeed, the federal courts lack the power to Rorder the States to prosecute individuals even when the failure to prosecute violatesthe Constitution. Cf. Linda R.S. v. Richard D., 410 U.S. 614, 619 & n.6 (1973) (citingGomez v. Perez, 409 U.S. 535 (1973)) (finding a lack of standing for a private citizento have criminal law enforced). Without such prosecutions there would of course beno case for the Supreme Court to review.139 For an argument that a treaty should be construed to grant the President the

power to order the stay of an execution of a State prisoner when the InternationalCourt of Justice (ICJ) orders such a stay as a provisional measure, see Carlos ManuelVazquez, Breard and the Federal Power to Require Compliance with ICJ Orders of ProvisionalMeasures, 92 AM. J. INT’L L. 683 (1998). See also Vazquez, Less than Zero?, supra note86, at 564–65 (discussing the President’s authority to require compliance with a final Rjudgment of the ICJ). But see Medellın v. Texas, 552 U.S. 491, 527 (2008) (holdingthat the President lacked power to order Texas to comply with ICJ final judgment).140 See Bollman, 8 U.S. (4 Cranch) at 93–94. But cf. Boumediene v. Bush, 553 U.S.

723, 744–45 (2008) (finding congressional preclusion of habeas jurisdiction for per-sons held in executive detention unconstitutional). After the McLeod incident, Con-gress gave the lower federal courts the power to grant habeas relief to personsrestrained of their liberty by federal or State officials in violation of the law of nations.Act of Aug. 29, 1842, ch. 257, 5 Stat. 539 (codified as amended at 28 U.S.C.§ 2241(c)(4) (2006)). See generally Bederman, supra note 134, at 527–31 (presenting Ra history of the grant and the controversies behind it).



state branch of customary international law had the force of preemp-tive federal law and that it had the status of general common law.

Of the five cases typically cited for the proposition that Statecourt interpretations of customary international law were understoodnot to be reviewable by the Supreme Court,141 Huntington v. Attrill142

is inapposite because it involved the law merchant.143 Although thelaw merchant was once considered a branch of the law of nations, itdiffered from the state-to-state branch of the law of nations in that itwas recognized to be subject to local variation.144 Today, the lawmerchant, or general commercial law, is not regarded as a part of cus-tomary international law. Adherents of the modern position do notcontend that this law is preemptive of State law.

A second case cited by revisionists as establishing the lack ofSupreme Court appellate jurisdiction over customary internationallaw, Oliver American Trading Co. v. Mexico,145 is inapposite because itinvolved an entirely distinct question. The issue in Oliver AmericanTrading was whether the district court had properly certified a ques-tion of foreign sovereign immunity for direct review by the SupremeCourt. Section 238 of the Judicial Code authorized the direct reviewby writ of error of questions of jurisdiction,146 which the Court con-strued to authorize review of “the question of jurisdiction of the dis-trict court as a federal court.”147 Although the Court did describe theimmunity question as one of “general law,” its decision was not basedon the view that a denial of this immunity by the State courts wouldnot be reviewable in the Supreme Court. Rather, the Court held thatthe certificate was not proper because the district court’s jurisdiction

141 The five cases discussed here are the ones cited by Bradley and Goldsmith. SeeBradley & Goldsmith, Current Illegitimacy, supra note 9, at 331 & n.64; Bradley & Gold- Rsmith, Critique, supra note 9, at 824 & n.48; Bradley, Goldsmith & Moore, supra note R33, at 913 & n.232; see also Ku, State Courts, supra note 16, at 277 n.52; Ku & Yoo, supra Rnote 16, at 202–03 & nn.182–84. R

142 146 U.S. 657 (1892).143 See id. at 673–74; see also Roth v. Ehman, 107 U.S. 319, 319 (1883) (granting

motion to dismiss because the Court had no jurisdiction to review a State court rulingon the validity of a foreign marriage), cited in Attrill, 146 U.S. at 683.144 See Bellia & Clark, supra note 21, at 20–22. R

145 264 U.S. 440 (1924).146 Judicial Code of 1911, Pub. L. No. 61-475, § 238, 36 Stat. 1087, 1157, amended

by Act of Jan. 28, 1915, Pub. L. No. 63-241, § 2, 38 Stat. 803, 804 (“Appeals and writs oferror may be taken from the district courts . . . direct to the Supreme Court in thefollowing cases: In any case in which the jurisdiction of the court is in issue, in whichcase the question of jurisdiction alone shall be certified to the Supreme Court fromthe court below for decision . . . .”).147 Oliver Am. Trading Co., 264 U.S. at 442 (emphasis added).



as a federal court was not based on the presence of the question ofimmunity. Federal jurisdiction in the case was based on diversity ofcitizenship.148 According to the Supreme Court, the sovereign immu-nity question did not concern the federal court’s jurisdiction as a fed-eral court, but was a question “of general law applicable alike toactions brought in other tribunals.”149 Oliver American Trading thusheld that jurisdiction was lacking because federal jurisdiction was notpremised on Mexico’s sovereign immunity defense, not because thesovereign immunity defense did not arise under federal law.150 In thislight, the most relevant point of Justice Brandeis’s opinion for theCourt is one that cuts in favor of the modern position—namely, itsrecognition that the immunity question was “applicable as fully to suitsin the [S]tate courts as to those prosecuted in the courts of the UnitedStates.”151

148 See Transcript of Record at 5–6, Oliver Am. Trading Co., 264 U.S. 440 (No. 662)(Removal Petition) (noting that the amount in dispute exceeds $3000 and that thedefendant is an alien and a nonresident of New York and the plaintiff is a resident ofDelaware); id. at 78 (Opinion Dismissing for Lack of Jurisdiction) (“On November 11,1922 the Mexican Government . . . and the National Railways of Mexico (GovernmentAdministration) . . . removed the action to this court, it being alleged that the suit wasbetween plaintiff, a Delaware corporation, and aliens, to wit: The Government of theUnited States of Mexico, a sovereign State, and National Railways of Mexico, a corpo-ration organized under the laws of that country.”).149 Oliver Am. Trading Co., 264 U.S. at 442. The Court may have regarded foreign

sovereign immunity as an immunity from jurisdiction over the person of the foreignstate, cf. Caleb Nelson, Sovereign Immunity as a Doctrine of Personal Jurisdiction, 115HARV. L. REV. 1559 (2002) (arguing that sovereign immunity bears on personal juris-diction), or it may have regarded it as an immunity from liability, cf. Vazquez, supranote 53, at 1084 n.132 (arguing that foreign sovereign immunity is in part an immu- Rnity from liability). When it enacted the FSIA in 1976, Congress made the absence ofsovereign immunity a condition of the subject-matter jurisdiction of the federal courtsover suits against foreign states. See 28 U.S.C. § 1330 (2006).150 Had Mexico sought to remove on the ground that its sovereign immunity

defense presented a federal question, its petition would have run afoul of the well-pleaded complaint rule. See supra note 63. R151 Oliver Am. Trading Co., 264 U.S. at 443. The subsequent decision in Transportes

Maritimos do Estado v. Almeida, 265 U.S. 104 (1924), cited by Weisburd for the proposi-tion that the question of immunity was not regarded as a federal question subject toreview in the federal courts, see Weisburd, supra note 18, at 40, actually confirms that ROliver American Trading held merely that “the claim of sovereign immunity does notpresent a question of federal jurisdiction within the meaning of § 238,” Almeida, 265U.S. at 105 (emphasis added). The Court in Almeida dismissed the writ of error onthe same ground. Id.; see also id. (distinguishing other cases that did involve the“question . . . of the jurisdiction of the court as a federal court”).



The third case relied upon by revisionists, Wulfsohn v. RussianSocialist Federated Soviet Republic,152 is a brief memorandum decisionthat raises more questions than it answers. The State court of appealsin Wulfsohn had dismissed claims against the Soviet Union on theground of sovereign immunity.153 The plaintiff in error appealed tothe Supreme Court raising constitutional objections, namely, that theNew York court’s decision according immunity to an unrecognizedgovernment contravened the Executive’s exclusive power over recog-nition of foreign sovereigns and that the dismissal of the suit was ataking of the plaintiff’s property without just compensation in viola-tion of the Fourteenth Amendment.154 The plaintiff in error did notargue that the Supreme Court had appellate jurisdiction because thequestion of sovereign immunity was one of federal law. The statutegoverning writs of error to the Supreme Court, Section 25 of the FirstJudiciary Act, extended jurisdiction (in relevant part) only to cases“where any title, right, privilege, or immunity is claimed under theconstitution, or any treaty or statute of or commission held or author-ity exercised under the United States” and the decision is against theparty claiming such right, title, privilege, or immunity.155 Because the Statecourt’s decision was in favor of the claimed right to immunity, theSupreme Court would have lacked jurisdiction by writ of errorwhether or not the defendant’s right to sovereign immunity had beenregarded as a matter of federal law. The Supreme Court summarilydismissed for lack of jurisdiction.156 The Court did not explain why itbelieved it lacked jurisdiction over the plaintiff in error’s constitu-tional claims. Presumably the Court agreed with the defendant inerror’s argument that the State court decision rested on independentand adequate state grounds.157 It is true that the defendant in error

152 266 U.S. 580 (1924) (per curiam), cited in Ku, State Courts, supra note 16, at R318–19, 335 & n.338, and Weisburd, supra note 18, at 39–40. R

153 Wulfsohn v. Russian Socialist Federated Soviet Republic, 138 N.E. 24, 26 (N.Y.1923).154 Brief in Behalf of Plaintiffs in Error at 4–5 Wulfsohn, 266 U.S. 580 (No. 65).155 Act of Sept. 6, 1916, ch. 448, § 2, 39 Stat. 726, 727 (1916). The plaintiff in

error could have sought a (discretionary) writ of certiorari asking the Court to reviewthe State court’s immunity decision, as the Supreme Court had recently been giventhe power to review by certiorari State court decisions upholding claims of federalright, see id., but he instead pursued his appeal as of right through writ of error.156 The opinion of the Supreme Court provides in its entirety as follows: “Dis-

missed for the want of jurisdiction upon the authority of the Act of September 6,1916, c. 448, § 2, 39 Stat. 726; Oliver American Trading Co., v. Mexico, 264 U.S. 440.”Wulfsohn, 266 U.S. at 580.157 See Motion to Dismiss for Lack of Jurisdiction or to Affirm for Want of Substan-

tial Issues at 6–7, Wulfsohn, 266 U.S. 580 (No. 65).



also cited Oliver American Trading for the proposition that theSupreme Court lacks jurisdiction over the immunity issue because it isa matter of “general law.”158 The Supreme Court’s citation of OliverAmerican Trading may suggest that it was endorsing this over-reading ofOliver American Trading. But, even if that was the Court’s intention,the cursory memorandum opinion in Wulfsohn is a slim reed on whichto base the claim that the Supreme Court was thought to lack thepower to review State court decisions interpreting customary interna-tional law, given that appellate jurisdiction over the sovereign immu-nity issue was lacking anyway and that such a reading would in anyevent fail to explain why the Court lacked jurisdiction over the consti-tutional claims.

The fourth case cited by revisionists, New York Life Insurance Co. v.Hendren,159 is similarly vague. Hendren involved a writ of error toreview the effect of the outbreak of the Civil War on the obligations ofa private insurance contract between a New York insurance companyand a citizen of Virginia. The Court concluded that it lacked jurisdic-tion,160 and revisionists argue that the Court did so because itregarded the laws of war as a matter of general law rather than one offederal law.161 This reading of Hendren derives some surface plausibil-ity from the fact that Justice Bradley, in dissent, put forth a strongstructural argument for treating the state-to-state branch of the law ofnations as preemptive federal law reviewable in the Supreme Court.Bradley wrote:

[T]he laws which the citizens of the United States are to obey inregard to intercourse with a nation or people with which they are atwar are laws of the United States. These laws will be the unwritteninternational law, if nothing be adopted or announced to the con-trary; or the express regulations of the government, when it sees fitto make them. But in both cases it is the law of the United States forthe time being, whether written or unwritten.162

In Bradley’s view, the case fell within the Supreme Court’s appellatejurisdiction within the meaning of Section 25, for reasons similar tothose set forth in the cases discussed by Bellia and Clark: “The powergiven by the Constitution to Congress to declare war, and the author-

158 Id. at 5, 9–10, 12.159 92 U.S. 286 (1875).160 See id. at 286–87.161 Bradley & Goldsmith, Critique, supra note 9, at 824 & n.53; Weisburd, supra R

note 18, at 39. R

162 Hendren, 92 U.S. at 288 (Bradley, J., dissenting).



ity of the general government in carrying on the same, are thegrounds on which the exemption or immunity is claimed.”163

Bradley’s opinion in Hendren shows that one of the contendingpre-Erie positions was that the state-to-state portion of customary inter-national law had the status of federal law. The fact that Bradley was alone dissenter, on the other hand, does not necessarily show that themajority was rejecting the structural constitutional case for treatingthis portion of the law of nations as preemptive of State law. Themajority may have believed instead that the structural argument fortreating the laws of war as preemptive was inapplicable to a “sectionalcivil war.”164 Moreover, the majority’s holding that jurisdiction waslacking appears to have been based on a reading of Section 25 of theFirst Judiciary Act that is no longer good law. As noted, Section 25 atthe time extended jurisdiction (in relevant part) only to cases “whereany title, right, privilege, or immunity is claimed under the constitu-tion, or any treaty or statute of or commission held or authority exer-cised under the United States.”165 The majority concluded that itlacked jurisdiction because it “nowhere appear[ed] that the constitu-tion, laws, treaties, or executive proclamations, of the United Stateswere necessarily involved in the decision” of the State court.166 Themajority in Hendren thus may have understood the general law of warto be binding on the States and potentially reviewable in the SupremeCourt, yet not to be “the constitution” or a “treaty or statute . . . orcommission” within the meaning of Section 25. To the extent theCourt was resting on the wording of Section 25, its interpretation ofthe statute is no longer good law, as the Supreme Court has subse-quently made clear that State court applications of federal commonlaw are reviewable by the Supreme Court, despite the fact that § 1257(the successor to Section 25) continues to refer to cases presenting

163 Id.164 Id. at 286 (majority opinion); cf. Caperton v. Bowyer, 81 U.S. (14 Wall.) 216,

228–29 (1871) (argument of counsel for defendant in error) (“It is true that thecourts of the United States, like the courts of the States, and of all other civilizedcountries, recognize the law of nations as binding upon them; and it is argued that asthe government of the United States is charged with the management and control ofour foreign relations, the courts of the United States ought to have the power ofdeciding in the last resort, all questions of international law, otherwise difficulties mayarise with foreign nations on account of erroneous decisions by the State courts whichthe government of the United States could not provide against. But whatever forcethe argument of convenience might have in a case arising between the United Statesand a foreign nation, it has very little in a case arising out of a defunct rebellion.”).165 See Judiciary Act of 1789, ch. 20, § 25, 1 Stat. 83, 85–87, amended by Act of Feb.

5, 1867, ch. 28, 14 Stat. 385.166 Hendren, 92 U.S. at 287.



rights or immunities under the Constitution, treaties, statutes, andcommissions of the United States.167 It is difficult to know the major-ity’s reasoning in Hendren, as the opinion contained very little analysisof the jurisdictional question.168

167 28 U.S.C. § 1257(a) (2006); see, e.g., Semtek Int’l Inc. v. Lockheed MartinCorp., 531 U.S. 497, 499–500, 508 (2001). I recognize, of course, that, “the modernconception of federal common law . . . did not exist circa 1788,” or, indeed, beforeErie. See Henry Paul Monaghan, Supremacy Clause Textualism, 110 COLUM. L. REV. 731,741 (2010). My point is that the Court’s reading of Section 25 in Hendren as permit-ting review only of the specified types of federal law has since been rejected. TheCourt’s literal interpretation of Section 25 in Hendren pretermitted any question ofwhether the law of nations was binding on the States or whether State court applica-tions of such law were reviewable in the Supreme Court as a constitutional matter.

The decision in City & County of San Francisco v. Scott, 111 U.S. 768 (1884), citedby Weisburd, supra note 18, at 39 & nn.235–37, may have rested on the same statutory Rground. See Scott, 111 U.S. at 769 (“[The question submitted to the Court] does notdepend on any legislation of Congress, or on the terms of the treaty . . . .” (citingHendren, 92 U.S. 286). The precise nature of the issue of “general public law” that theCourt was asked to review in Scott is unclear from the opinion, so it is difficult to saywhether it implicated the state-to state portion of the law of nations. Cf. infra note168 (noting that the “law of war” issue in Hendren may not have been part of the state- Rto-state portion of the law of nations).168 It is worth noting, additionally, that the legal principle involved in Hendren

likely did not belong to the portion of the law of nations that corresponds to today’spublic international law. It is true that, in 1875, the laws of war were understood toinclude a rule suspending private contracts between citizens of belligerent states. See2 WILLIAM WINTHROP, MILITARY LAW AND PRECEDENTS 1209 (Boston, Little, Brown, &Co. 2d rev. ed. 1896) (1886) (stating that the law of war creates a strict rule ofnonintercourse between the populations of the warring parties, and that “existingcontracts and pecuniary obligations are suspended”). But, as noted, the law ofnations was then thought to encompass areas of the law that we do not regard as partof public international law today. It appears that the effect of war on a private con-tract is today regarded as within the province of domestic contract law, not the laws ofwar. Thus, in a 1959 treatise on the modern law of war, the chapter on “IntercourseBetween Belligerents” discusses intercourse between the belligerent governments andarmies. MORRIS GREENSPAN, THE MODERN LAW OF LAND WARFARE 376–99 (1959).With respect to individuals, it says merely that, as a matter of national law, “the con-tending states usually forbid as a serious crime unauthorized communications, eventhough innocent in nature, between persons under their control and those in enemyterritory.” Id. at 378 (citing Trading with the Enemy Act, 12 U.S.C. § 95a (2006);Trading with the Enemy Act, 1939, c. 39 (Eng.)). By contrast, the 1934 Restatement ofContracts did address the effect of war on a private contract, specifying that “[w]arsupervening after the formation of contracts between citizens of the United Statesand citizens of the enemy country” has the effect of rendering the contract illegal if itsperformance “involves communication across the line of hostilities, or . . . will aid theenemy or diminish the power of the United States to carry on war effectively.”RESTATEMENT OF CONTRACTS § 596(b) (1934).



The final case often cited as establishing that the Supreme Courtwas thought to lack jurisdiction to review the State courts’ applicationof customary international law is Ker v. Illinois,169 which presented thequestion whether the State court’s jurisdiction in a criminal case wasvitiated by the fact that the defendant had been brought to the juris-diction against his will from Peru by a private party.170 The defendantraised arguments based on the Due Process Clause, a federal statute,and a treaty, but not customary international law. After rejecting eachof the defendant’s claims, the Court commented that the defendantlikely had no claim under customary international law either, but that,even if he had had one, the issue would not be within the SupremeCourt’s jurisdiction to review.171 As in Hendren, that proposition(stated without citation in Ker) might have been based on the ideathat, even if the defendant enjoyed immunity under customary inter-national law, the immunity was not conferred by the Constitution or astatute, treaty, or commission within the meaning of Section 25. Inany event, the precedential force of this statement is diminished bythe fact that Ker did not rely on customary international law at all, andthe Court thought that such a claim would have been frivolous in anyevent.172 More noteworthy is the Court’s statement that the Statecourts are “bound to take notice” of customary international law.173

This dictum in Ker shows that, whatever the reason for its view thatappellate jurisdiction over the customary international law issue waslacking, the Court did not equate the lack of Supreme Court reviewwith a freedom of the States to disregard customary international law.

In sum, there is substantial support for the proposition that,before Erie, the portion of the law of nations that corresponded topresent-day customary international law was understood to be bindingon the States. The cases cited by revisionists for the claim that Stateswere free to violate such law do not establish that point. As for the

169 119 U.S. 436 (1886).170 Id. at 443 (“[I]t was a clear case of kidnapping within the dominions of Peru,

without any pretence of authority . . . from the government of the United States.”).171 See id. at 444.172 The Court may have believed that a private abduction does not violate interna-

tional law. But cf. Bellia & Clark, supra note 41, at 448–49 (noting that in 1789, inju- Rries to aliens by U.S. citizens were thought to violate the law of nations if notremedied). Or the Court may have believed that the law of nations required theextradition or punishment of the abductor or the payment of damages but not therelinquishment of jurisdiction. But cf. Carlos Manuel Vazquez, Treaty-Based Rights andRemedies of Individuals, 92 COLUM. L. REV. 1082, 1158 (1992) (arguing that moderninternational law requires the repatriation of a persons abducted from foreign terri-tory in violation of international law).173 Ker, 119 U.S. at 444.



Supreme Court’s jurisdiction to review State court applications of cus-tomary international law, the cases are equivocal at best insofar as theyconcern public international law norms, and may just have rested on areading of the statute regulating the Supreme Court’s appellate juris-diction that has since been rejected.174 In any event, these opinionsdid not equate lack of reviewability with State freedom to depart fromcustomary international law. Even the cases that most strongly sup-port the claim that customary international law was not reviewable bythe Supreme Court insisted that the States were bound by such law,even though States were understood to have the power to depart fromother aspects of the general law. Though there appear to be no fed-eral cases invalidating State laws on the ground of conflict with cus-tomary international law, this appears to be a result of the absence ofany such State laws, as there also appear to be no federal cases uphold-ing State laws that purport to alter or depart from customary interna-tional law.

c. The General Law as an Intermediate Status

The pre-Erie doctrinal landscape strongly supports the claim thatcustomary international law was understood to be binding on theStates, whether or not the State courts’ applications of such law werereviewable in the Supreme Court. Even if such law had been viewedas “general law” in all respects, however, the pre-Erie approach to suchlaw would not provide much of a historical antecedent for treatingcustomary international law as modern-day State law. Given the mind-set of judges of that era, the pre-Erie general law regime, as it oper-ated, resembled present-day federal law more than present-day Statelaw.

The Founders sought to ensure that cases possibly implicating thelaw of nations would be decided in the federal courts, and so theyextended the federal judicial power to cases involving ambassadors,foreign states, aliens, and admiralty.175 The Judiciary Act of 1789, in

174 See supra note 167 and accompanying text. R

175 See supra note 126; see also THE FEDERALIST NO. 3, at 43 (John Jay) (Clinton RRossiter ed., 1999) (“Under the national government, treaties [and] the laws ofnations, will always be expounded in one sense and executed in the same manner—whereas adjudications on the same points and questions in thirteen States, or in threeor four confederacies, will not always accord or be consistent; and that, as well fromthe variety of independent courts and judges appointed by different and independentgovernments as from the different local laws and interests which may affect and influ-ence them.”); THE FEDERALIST NO. 99, at 477 (Alexander Hamilton) (Clinton Rossiter Red., 1961) (“So great a proportion of the cases in which foreigners are parties involve



turn, granted federal courts jurisdiction in these cases,176 thus dimin-ishing the need for Supreme Court review of these questions from theState courts. The lower federal courts, for their part, used their ownindependent judgment regarding the content of this law.177 Moreo-ver, even when State courts decided issues of customary internationallaw, they could be expected to arrive at interpretations that accordedwith those of the federal courts. During this era, federal and Statecourts alike understood that, in applying the general common law,they were discovering a set of norms that had an existence apart fromwhat the courts said about them. As Professor (now-Judge) Fletcherhas documented, because of their sense of common enterprise, fed-eral and State courts frequently relied on one another’s decisions incases involving the general common law, and the result was a remarka-ble degree of uniformity in the interpretation of this law.178 The pos-sible threat to this uniformity posed by the State legislatures’recognized power to depart from the common law never materializedwith respect to the state-to-state portion of the law of nations, perhapsbecause the States did not understand themselves to be free to departfrom the latter.

Post-Erie common law, however, exhibits none of these character-istics. States understand that they have the power to depart from theapproach of other States and of the federal courts, not just legisla-tively, but also through judicial decision. The result has been a greatdiversity of approaches to common law matters among the States.This diversity has not extended to customary international law, pre-sumably because of the “well-entrenched” view that such law is fed-eral.179 Were the revisionist approach to be adopted, however, theStates would know that they are free to adhere to such law or not, or

national questions that it is by far most safe and most expedient to refer all those inwhich they are concerned to the national tribunals.”).176 Judiciary Act of 1789, ch. 20, § 9, 1 Stat. 73, 76–77 (granting district courts

exclusive original jurisdiction over admiralty and maritime cases and concurrent juris-diction over alien tort suits); id. § 11, 1 Stat. at 78 (granting circuit courts concurrentjurisdiction over suits in diversity); id. § 13, 1 Stat. at 80–81 (granting the SupremeCourt exclusive jurisdiction over cases brought against ambassadors and original, butnot exclusive, jurisdiction over suits brought by ambassadors).177 See, e.g., Huntington v. Attrill, 146 U.S. 657, 683 (1892) (“[I]t is one of those

questions of general jurisprudence which [the federal] court must decide for itself,uncontrolled by local decisions.”).178 See Fletcher, supra note 124, at 1554, 1562–64, 1572–75. Fletcher’s analysis R

focused on the law merchant, but his analysis is a fortiori applicable to the state-to-statebranch of the law of nations, which was understood to be binding on States as a mat-ter of international law.179 Bradley & Goldsmith, Critique, supra note 9, at 816. R



to depart or alter it as they wish or as instructed by their legislatures ortheir citizens through referenda such as Oklahoma’s. Such anapproach to customary international law would be radically differentfrom the approach that prevailed before Erie, whether its status thenwas federal law or general law.

3. Post-Erie Doctrine

The post-Erie decision most often cited by proponents of themodern position is Banco Nacional de Cuba v. Sabbatino.180 As revision-ists often note, relying on Sabbatino as support for the modern posi-tion is ironic, since the Court’s holding was that customaryinternational law could not be applied to question the validity of cer-tain acts of foreign states.181 As discussed below, Sabbatino actuallyshows that not all of customary international law is federal law for allpurposes.182 But the portion of the Court’s opinion holding that theact-of-state doctrine itself has the status of federal law does stronglysupport the modern position, albeit in dictum.183

The Court in Sabbatino explained that the act-of-state doctrine—which generally prohibits courts from questioning the validity of anact of a foreign state within its own territory—is based on separation-of-powers principles. The Constitution assigns to the President theresponsibility for conducting the nation’s foreign relations, and thecourts’ failure to give effect to certain acts of state would hinder theperformance of this function. Because the State courts’ failure to giveeffect to such acts would be just as harmful as the federal courts’ fail-ure to do so, the Court held, the act-of-state doctrine is equally appli-cable in State courts.184 In reaching the conclusion that the act-of-state doctrine thus preempted any contrary State law or policy, theCourt relied on an article by Professor (later Judge) Philip Jessup,written shortly after Erie was decided. Jessup addressed the applicabil-ity of Erie to customary international law and concluded that Erie did

180 376 U.S. 398 (1964).181 Bradley & Goldsmith, Critique, supra note 9, at 860; Young, supra note 19, at R

440–41.182 See infra text accompanying notes 208–22. R

183 Bellia and Clark argue that Sabbatino directly supports their claim that a subsetof customary international law is preemptive federal law. I consider this claim in sub-part II.D, below.184 The Court left open whether State courts could decide to give effect to foreign

acts of state even if the act-of-state doctrine did not require them to do so, Sabbatino,376 U.S. at 425 n.23, but it held that State courts must give effect to such acts whenthe act-of-state doctrine applies.



not require that customary international law be treated as State law.185

The Supreme Court agreed with his analysis and found it equallyapplicable to the act-of-state doctrine:

It seems fair to assume that the Court did not have rules like the actof state doctrine in mind when it decided Erie R. Co. v. Tompkins.Soon thereafter, Professor Philip C. Jessup, now a judge of the Inter-national Court of Justice, recognized the potential dangers wereErie extended to legal problems affecting international relations.He cautioned that rules of international law should not be left todivergent and perhaps parochial [S]tate interpretations. His basicrationale is equally applicable to the act of state doctrine.186

The Court’s endorsement of Jessup’s view that customary interna-tional law is federal law is admittedly dictum, as the Court noted thatthe act-of-state doctrine was not itself required by international law.187

But it was well-considered dictum.188 Indeed, the Court seemed toregard Jessup’s view concerning customary international law and itsown view concerning the act-of-state doctrine as instantiations of abroader principle that it appeared to regard as self-evident: that Eriewas inapplicable “to legal problems affecting international rela-tions.”189 The Restatement was thus well justified in regarding Sabbatinoas strong support for the modern position.190

Revisionists, for their part, claim Chief Judge Hand’s opinion inBergman v. De Sieyes191 as support.192 In this post-Erie, pre-Sabbatinocase, Hand stated that New York law was applicable to the question of

185 See Philip C. Jessup, The Doctrine of Erie Railroad v. Tompkins Applied to Interna-tional Law, 33 AM. J. INT’L L. 740, 743 (1939).186 Sabbatino, 376 U.S. at 425 (footnote omitted).187 Id. at 421.188 The Court itself noted that it could have avoided the Erie question, since New

York law did not conflict with federal law, but it felt “constrained to make it clear thatan issue concerned with a basic choice regarding the competence and function of theJudiciary and the National Executive in ordering our relationships with other mem-bers of the international community must be treated exclusively as an aspect of fed-eral law.” Id. at 425.189 See id.; cf. infra text accompanying notes 507–508 (arguing that preemptive R

force of customary international law may be regarded as a doctrinal manifestation of abroader preemption of State action that unduly interferes with the federal govern-ment’s conduct of foreign relations).190 RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW § 111 n.3 (1987).191 170 F.2d 360 (2d Cir. 1948).192 Young claims that Bergman is an “unequivocal” rejection of the modern posi-

tion. Young, supra note 19, at 458 n.479; see also Bradley & Goldsmith, Critique, supra Rnote 9, at 828 (noting that Chief Judge Hand “reached a conclusion contrary to Jes- Rsup’s”); Ku & Yoo, supra note 16, at 203 & n.186 (arguing that Chief Judge Hand R“treated [customary international law] as part of New York’s common law”).



a foreign diplomat’s immunity, an issue addressed by customary inter-national law. But Hand actually left open whether State law wouldapply if it conflicted with international law. Indeed, Hand acknowl-edged that “an avowed refusal to accept a well-established doctrine ofinternational law, or a plain misapprehension of it, [might] present afederal question.”193 He found it unnecessary to reach this issue,since New York law was consistent with international law.194 Thus,Hand went no further than to hold that State law applies to the extentthat it is consistent with customary international law, which is justanother way to put the modern position.

More relevant is the very recent decision in Samantar v. Yousuf,195

which shows that virtually no one today embraces the view that revi-sionists attribute to Hand. The issue in Samantar was whether theimmunity of foreign officials was governed by the Foreign SovereignImmunities Act (FSIA). The executive branch’s amicus brief arguedthat the immunity was not covered by the FSIA, but that foreign offi-cials were entitled to a “common law” immunity.196 Although thebrief does not call it a federal common law immunity, it is clear that theexecutive branch does not regard it as having the force of State law, asits brief maintains that the content of this immunity is determined by“principles adopted by the executive branch, informed by customaryinternational law.”197 The petitioner, for his part, argued that theFSIA governs the immunity of foreign officials for official conductbecause customary international law entitles them to such an immu-nity,198 a position that Bradley and Goldsmith had advanced.199 As afall back, however, petitioners would have recognized a common lawimmunity.200 Finally, respondents also recognized that foreign offi-cials would be covered by a common law immunity, although they

193 See Bergman, 170 F.2d at 361.194 See id.195 130 S. Ct. 2278 (2010).196 See Brief for the United States as Amicus Curiae Supporting Affirmance at 8,

Samantar, 130 S. Ct. 2278 (No. 08-1555).197 See id. at 8; see also Statement of Interest of the United States of America at 5–6,

Yousuf v. Samantar, No. 1:04 CV 1360 (E.D. Va. Feb. 14, 2011) (arguing that “theExecutive Branch . . . play[s] the primary role in determining the immunity of foreignofficials as an aspect of the President’s responsibility for the conduct of foreign rela-tions and recognition of foreign governments”).198 Brief of Petitioner at 43–44, Samantar, 130 S. Ct. 2278 (No. 08-1555). They

argue that the statute should be read to incorporate this international law immunity.Id. at 24–26, 32–34.199 Curtis A. Bradley & Jack L. Goldsmith, Foreign Sovereign Immunity, Individual

Officials, and Human Rights Litigation, 13 GREEN BAG 2D 9, 16 (2009).200 Brief of Petitioner, supra note 198, at 24–34. R



would have tied it more directly to customary international law andurged a less prominent role for the executive branch.201 Again, thereis no suggestion that the contemplated “common law” immunitydepended in any way on State law.

The Court ultimately rejected the petitioners’ argument that theimmunity of foreign officials was governed by the FSIA, holding thatthe immunity of foreign officials is ordinarily “properly governed bythe common law,” as it was for foreign states before the enactment ofthe FSIA.202 Although the Court did not specify the nature of thiscommon law, the Court’s discussion of the pre-FSIA regime leaves nodoubt that it regarded the relevant law as federal, not State, law.203

Thus, neither the Court nor any of the parties in Samantar embracedthe revisionist position that the immunity of foreign officials is gov-erned by State law.204 The Court’s characterization of the immunityas one under the “common law” and its description of the Executive’spre-FSIA role may suggest that it understood customary internationallaw to be applicable in this context only as filtered through the execu-tive branch. As discussed further below,205 however, the opinion isprobably best read to leave open this and all other questions about thenonstatutory immunity of foreign officials apart from its federalnature.

B. The Limits of the Modern Position

Contrary to the revisionists’ claims, neither the Restatement normost of its defenders have maintained that all norms of customaryinternational law have the status of preemptive federal law applicablein the courts in all circumstances in which they address the matter indispute. The Restatement recognizes that norms of customary interna-tional law sometimes lack the status of federal law in the face of cer-tain acts of foreign states and of the President and possibly otherexecutive branch officials. This subpart examines these established

201 Brief for the Respondents at 34–35, 53–55, Samantar, 130 S. Ct. 2278 (No. 08-1555); Transcript of Oral Argument at 35–36, 38–39, 42–44, Samantar, 130 S. Ct. 2278(No. 08-1555).202 Samantar, 130 S. Ct. at 2292.203 See id. at 2284 (noting that, pre-FSIA, if the State Department had not

expressed a view regarding immunity, “a district court inquired ‘whether the groundof immunity is one which it is the established policy of the [State Department] torecognize’” (alteration in original) (quoting Republic of Mexico v. Hoffman, 324 U.S.30, 36 (1995))).204 See Bradley & Goldsmith, Critique, supra note 9, at 828 (citing Bergman v. De R

Sieyes, 170 F.2d 360 (2d Cir. 1948), with approval for this proposition).205 See infra text accompanying notes 550–557. R



limits of the modern position’s claim that customary international lawhas the force of federal law, and explains their compatibility with thestructural argument for regarding customary international law as pre-emptive federal law.

1. Sabbatino and the Inapplicability of Some CustomaryInternational Law Norms to Some Acts of Foreign States

In attacking the modern position, the revisionists describe it asmaintaining that all of customary international law has the force ofpreemptive federal law always and in all contexts.206 It is true that theThird Restatement states broadly, in its black letter, that “[i]nternationallaw . . . [is] law of the United States and supreme over the law of theseveral States” and that “[c]ourts in the United States are bound togive effect to international law.”207 But this seemingly unlimited viewis qualified in the comments and reporter’s notes, especially in theirdiscussion of Sabbatino. As noted above, revisionists often note theirony of the Restatement’s reliance on Sabbatino, given Sabbatino’s hold-ing that some rules of customary international law are not enforceablein federal or State courts in certain circumstances.208 But such reli-ance is not so odd if the modern position does not in fact insist that allcustomary international law is always enforceable as preemptive fed-eral law. The Restatement does not deny that Sabbatino precludes theapplication of some rules of customary international law in some con-texts,209 which in turn shows that the Restatement’s version of the mod-ern position is not as uncompromising as the revisionists claim.210

Sabbatino held that federal and State courts alike had to giveeffect to the act of a foreign state performed within its own territoryeven though it conflicted with the claimed rule of customary interna-tional law prohibiting discriminatory or uncompensated takings of theproperty of foreign nationals. The Court concluded that judicialenforcement of that rule in that context would interfere with theExecutive’s pursuit of the nation’s foreign policy interests. In reach-ing this conclusion, the Court emphasized the disputed and contro-versial character of the rule. Capital-importing states at the timevociferously objected to the rule, and as a result the rule “touch[ed]

206 Bradley & Goldsmith, Critique, supra note 9, at 846 & n.202; Young, supra note R19, at 377–78. R207 RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW § 111(1), (3) (1987).208 See supra text accompanying note 181. R209 See RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW § 111 n.1.210 Some scholarly defenders of the modern position are quite explicit in noting

that it does not insist that all customary international law is always preemptive federallaw. See Goodman & Jinks, supra note 16, at 480; Neuman, supra note 9, at 376. R



. . . sharply on national nerves.”211 The holding of Sabbatino might begeneralized as precluding the application of rules of customary inter-national law to question the validity of an act of a foreign state withinits own territory if such rules are disputed and touch sharply onnational nerves. In other words, rules of customary international lawpossessing such characteristics lack the status of federal law in thatparticular context.

One possible way to reconcile Sabbatino with the Restatement’sblack-letter rule that customary international law has the status of fed-eral law is to understand the Court to have held that there was in factno norm of international law prohibiting the discriminatory oruncompensated taking of foreign nationals’ property. The Court’sreliance on the disputed nature of the asserted international lawnorm might support such a reading.212 If a large enough number ofstates disputed the norm, the norm might lack the necessary statepractice and opinio juris to be a rule of customary international law.But the Court’s opinion is not susceptible to such a reading. Had theCourt denied the existence of the claimed norm, the Court wouldhave been taking a position contrary to the longstanding position ofthe executive branch on the matter.213 Yet, in explaining why thisnorm could not be applied in this context, the Court gave as one rea-son the interference with the Executive’s conduct of foreign relationsthat would result if the courts were to disagree with the Executive’sview.214 The Court’s reasoning here indicates that the Court was notin fact rejecting the Executive’s view that there was a rule of customaryinternational law prohibiting discriminatory or uncompensated tak-ings of foreign nationals’ property. Instead, the Court held that someclaimed rules of customary international law (namely, those that aredisputed and touch sharply on national nerves) may not be used by

211 Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398, 428 (1964).212 Dean Koh appears to read Sabbatino this way. Koh, supra note 16, at 1834–35. R

213 See Sabbatino, 376 U.S. at 429 & n.29.214 Id. at 431–33. Note that the Court here did not even consider the possibility

that the role of the courts was just to apply whatever interpretation of customary inter-national law the Executive was embracing at the moment. The Court appears to havebelieved that the courts’ role, if they did reach the international law issue, was toapply their own best judgment of what international law required. It was willing tofind the international norm inapplicable on allocation-of-powers grounds, but it wasunwilling to compromise the courts’ role as an independent judge of internationallaw if the question were in fact before it. Cf. infra notes 538–557 and accompanying Rtext (discussing the Court’s pre-FSIA approach to the immunity of foreign states andheads of state).



courts to invalidate the act of a foreign state within its own territoryeven if they are in fact rules of international law.215

The Restatement takes a different approach to reconciling Sabba-tino with its black-letter rule that “international law is law of theUnited States.” Its reporter’s notes explain that the black-letter rule isnot applicable in the Sabbatino context.216 The Restatement here citesan article by its Chief Reporter, Louis Henkin, which clarifies hisunderstanding of the relationship between the act-of-state doctrineand the idea that international law is part of our law. In a case likeSabbatino, he wrote,

the proposition that “international law is part of the law of theUnited States” . . . [is] irrelevant. Effectively, it means that thecourts will apply international law in an appropriate case against theUnited States, the one government which is subject to the law of theUnited States. In a case like Sabbatino, however, the Government ofCuba, acting in Cuba, obviously was not subject to the laws of theUnited States. . . . International law might have been relevant if itrequired the United States to respond to Cuba’s violation in a par-ticular way; for example, if international law forbade the UnitedStates to give effect to Cuba’s confiscations, American courts wouldcarry out that obligation and refuse to give them effect. But inter-national law does not tell the United States how to react to Cubanacts that violate international law. The United States is free to con-done, acquiesce in, implement, or even applaud them.217

This analysis contains an important insight (discussed further below),but the attempt to dissolve the apparent conflict between the act-of-state doctrine and the idea that international law is U.S. law by refer-ence to choice-of-law rules cannot succeed. Choice-of-law rules them-selves have the status of either State or federal law. Since the Court inSabbatino held that the act-of-state doctrine is a matter of federal law,presumably the doctrine’s applicability cannot turn on State choice-of-law rules. It follows that the applicability of the doctrine must turn onfederal choice-of-law rules. What are the relevant choice-of-law rules?Henkin appears to have had in mind territorial rules of the traditionalvariety. Thus, he stressed that, because Cuba was acting within its ownterritory, Cuban law would apply, and Cuban law did not incorporate

215 See Sabbatino, 376 U.S. at 428.216 RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW § 111 reporter’s note 1

(1987).217 Louis Henkin, Act of State Today: Recollections in Tranquility, 6 COLUM. J. TRANS-

NAT’L L. 175, 181 (1967) (footnote omitted); see also id. at 183 (making it clear thatHenkin viewed the applicability of act-of-state doctrine as determined by choice-of-lawprinciples).



the claimed rule of international law.218 But choice-of-law rules arenot set in stone, and the traditional territorial rules are not alwaysfollowed. In most cases there is nothing to prevent the United Statesfrom applying its own law or policy,219 which generally frowns upondiscriminatory or uncompensated takings of property,220 or fromdeclining to give effect to uncompensated takings by foreign statesthat violate international law. It is Sabbatino itself that adopts a federalrule barring the application of rules of international law having cer-tain characteristics to invalidate an act of a foreign state performedwithin its own territory. We might call this a federal choice-of-lawrule, but, in the end, it is a rule making certain rules of customaryinternational law inapplicable in our courts (State and federal) in thisparticular context, a rule that is at least in tension with the broadblack-letter rule that international law is U.S. law.

In the end, the Restatement recognizes that, in light of the act-of-state doctrine, customary international law is not enforceable in thecourts as federal law in certain circumstances. It gives an unpersuasiveexplanation based on the conflict of laws, but the source of the excep-tion is less important than the fact that the exception is recognized.The question, then, is which norms lack this status and in what cir-cumstances. As discussed above, Sabbatino is best read to hold thatdisputed norms that touch sharply on national nerves are inapplicablein the context of a suit challenging the act of a foreign state within itsown territory.

The conclusion that such norms are inapplicable in that contextis consistent with the allocation-of-powers rationale for according pre-emptive force to customary international law. As discussed above, thestructural concern is that violations of international law attributable tothe United States would interfere with the federal government’s abil-ity to conduct the nation’s foreign relations effectively. To avoidinterference with the federal government’s ability to achieve foreign

218 See id. at 181.219 The Constitution may limit the application of federal law if the case has no

connection with the United States. Cf. Allstate Ins. Co. v. Hague, 449 U.S. 302, 320(1981) (holding that the Due Process Clause of the Fourteenth Amendment limitsStates’ power to apply their laws to disputes having few contacts with the State). Butcf. Lea Brilmayer & Charles Norchi, Federal Extraterritoriality and Fifth Amendment DueProcess, 105 HARV. L. REV. 1217 (1992) (arguing that the Due Process Clause of theFifth Amendment incorporates different limits than the Due Process Clause of theFourteenth Amendment).220 U.S. CONST. amend. V (“[N]or shall private property be taken for public use,

without just compensation.”); Chi., Burlington & Quincy R.R. Co. v. Chicago, 166 U.S.226, 230 (1897) (holding that the Takings Clause of the Fifth Amendment applies tothe States through the Fourteenth Amendment).



relations goals, the courts will enforce international law to prevent vio-lations of international law by the States. But smooth and effectiveforeign relations do not require the courts to invalidate foreign actsthat violate international law. To the contrary, striking down such actsmight itself produce foreign relations problems. Thus, the allocation-of-powers rationale for according preemptive force to customary inter-national law supports Henkin’s view that this principle is relevant onlywhen international law is sought to be applied “against the UnitedStates” (including the States). The rule that Henkin characterized as achoice-of-law principle is better understood as stemming from theconstitutional structure, which compels the application of customaryinternational law against the United States and the States, but notagainst foreign states acting within their own territory. Territorialnotions of sovereignty, which underlie traditional choice-of-law rules,play a role in delineating the limits of the federal rule barring applica-tion of customary international law in this context. Presumably, thereason the constitutional structure does not forbid the invalidation offoreign state acts performed outside their territory is that, in light oftraditional notions of territorial sovereignty, the failure to give effectto such acts will not produce undue international friction. But thereal work in the analysis is done by the constitutional structure.

A second problem with Henkin’s choice-of-law explanation ofSabbatino’s holding is that it proves too much. The Court in Sabbatinodid not hold that customary international law can never be applied toinvalidate the acts of a foreign state within its own territory. Rather, itwas careful to limit its holding to norms of international law havingcertain characteristics. The implication of the Court’s analysis is thatnorms of customary international law that are not as disputed or donot touch as sharply on national nerves might be applied to invalidatethe act of a foreign state within its own territory. Under the territorialchoice-of-law rules applied by Henkin, however, Cuban law would beequally applicable, and U.S. law equally inapplicable, whether or notthe Cuban act of state violated clear and uncontroversial norms ofinternational law.221 On the other hand, the Court’s distinction can

221 Henkin did recognize that, under choice-of-law rules, the ordinarily applicablelaw could be displaced if it conflicted with the forum’s public policy. According toHenkin, the act-of-state doctrine is a special choice-of-law rule that operates to pre-clude courts (State and federal) from declining to give effect to the ordinarily applica-ble law because of a conflict with public policy. See id. at 178. Thus, perhaps Henkinwould have said that the act-of-state doctrine precluded resort to the public policyexception only if the rule of international law that was violated was disputed andtouched sharply on national nerves. But this characterization merely brings to thefore my earlier point that the real work in these cases is not being done by choice-of-



be explained in separation-of-powers terms. Even though the consti-tutional structure does not require the application of customary inter-national law to invalidate foreign acts of state in such contexts, theCourt could reasonably have concluded that the invalidation of for-eign acts that conflict with clear and uncontroversial norms of interna-tional law would not unduly interfere with the conduct of foreignrelations by the federal political branches, and that countervailingnotions of justice warrant the invalidation of such acts.222

2. The Paquete Habana and the Applicability of CustomaryInternational Law to Federal Officials

That the modern position does not insist that customary interna-tional law has the force of federal law always and for all purposes isalso shown by the recognition in The Paquete Habana that the applica-bility of such law is subject to “controlling executive . . . act[s].”223 Ifcustomary international law did have all of the attributes of federal lawfor all purposes, then it would follow that the President would havethe duty faithfully to execute it.224 That duty, however, would be inconflict with an acknowledgement that the President has the power toact in contravention of such law in at least certain circumstances.

Exactly in which circumstances the President may violate custom-ary international law has been the subject of heated debate.225 Someadherents of the modern position maintain that the President is freeto violate such law only when speaking for the United States at the

law rules, but rather by the separation-of-powers principles that justify a distinctionbetween accepted, undisputed norms and controversial, disputed ones.222 The Court does not tell us whether federal law requires the invalidation of for-

eign state acts that conflict with clear and uncontroversial norms of international lawor merely permits the invalidation of such acts if State law incorporates the relevantinternational law norms. Cf. Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398, 425n.23 (1964) (leaving open whether States may be more deferential to foreign statesthan the act-of-state doctrine requires). Leaving it to the States to decide the applica-bility of undisputed norms of customary international law to the acts of foreign stateswould not be inconsistent with the structural case for the modern position, althoughSosa appears to reflect a preference for treating the issue as a federal one. See infraPart I.C. If the issue were treated as a matter of State law, Supreme Court review ofState court decisions regarding the content of customary international law would haveto be available, for the reasons discussed infra Part IV.223 The Paquete Habana, 175 U.S. 677, 700 (1900).224 U.S. CONST. art. II, § 3.225 See, e.g., Agora: May the President Violate Customary International Law?, 80 AM. J.

INT’L L. 913 (1986); Agora: May the President Violate Customary International Law?(Cont’d), 81 AM. J. INT’L L. 371 (1987).



international plane as “sole organ” of the nation’s foreign relations.226

Others defend a broader presidential power to exempt executive offi-cials from complying with customary international law.227 Somecourts have held that cabinet-level officials have the authority to vio-late customary international law,228 a position that the Office of LegalCounsel has endorsed.229 On the other hand, it is widely understoodthat lower-level executive officials lack the power to violate customaryinternational law unless authorized by the President or perhaps a cabi-net-level official.230 In The Paquete Habana itself, the Court enforced arule of customary international law against a lower-level executiveofficial.

Exempting the President from certain norms of customary inter-national law would be consistent with the allocation-of-powers ratio-nale for according such norms the status of federal law. Theallocation-of-powers concern is that violations of such norms wouldinterfere with the federal government’s ability to conduct foreign rela-tions, and, of course, the Constitution allocates to the Presidentimportant foreign relations powers. As discussed below, Bellia andClark argue that the violation of certain norms of customary interna-tional law by the States would be in conflict with the federal power torecognize foreign governments, a power that is, in turn, inferred fromthe power to send and receive ambassadors. The power to receiveambassadors is allocated exclusively to the President; the power tosend ambassadors is shared by the President and the Senate.231 Tothe extent that the preemptive effect of such norms is traced to therecognition power, therefore, it would seem to follow that the Presi-dent himself is not bound and may even have the power to authorize

226 See Louis Henkin, The President and International Law, 80 AM. J. INT’L L. 930, 936(1986).227 See, e.g., Jonathan I. Charney, The Power of the Executive Branch of the United States

Government to Violate Customary International Law, 80 AM. J. INT’L L. 913, 921 (1986)(arguing that the President’s special role in the U.S. government gives him theauthority, acting alone, to place the United States in violation of customary interna-tional law).228 See, e.g., Garcia-Mir v. Meese, 788 F.2d 1446, 1453–55 (11th Cir. 1986) (finding

that the Attorney General had the authority to violate customary international law).229 Auth. of the Fed. Bureau of Investigation to Override Int’l Law in Extraterrito-

rial Law Enforcement Activities, 13 Op. O.L.C. 163, 163–64 (1989).230 Id. at 180; LOUIS HENKIN, FOREIGN AFFAIRS AND THE UNITED STATES CONSTITU-

TION 243–44 (2d ed. 1996).231 U.S. CONST. art. II, §§ 2–3.



violations.232 On the other hand, to the extent that a norm’s preemp-tive effect is traceable to the war power, one would think that the Pres-ident is bound, as the Constitution allocates the war power toCongress.233

Which norms exactly are binding on the President is beyond thescope of this Article. The present point is that even adherents of themodern position hold varying views on this question. What unitesthem is the claim that the customary international law binds theStates.

C. Sosa and the Modern Position

Revisionist scholars and at least one revisionist federal judge234

have read the decision in Sosa v. Alvarez-Machain as rejecting the mod-ern position. Some defenders of the modern position, on the otherhand, have read Sosa as definitively rejecting the revisionist view.235

(There is a great deal of truth in Young’s assertion that “Justice Sou-ter’s majority opinion in Sosa v. Alvarez-Machain has become some-thing of a Rorschach blot, in which each of the contending sides inthe debate over the domestic status of customary international law . . .sees what it was predisposed to see anyway.”236) In my view, Sosa doesnot definitively adopt either position, but its analysis strongly supportsthe modern one.

The issue in Sosa was whether a federal cause of action for dam-ages was available for the defendant’s alleged violations of customaryinternational law.237 The plaintiff had relied on § 1350 as creating acause of action,238 but the Court concluded that the provision waspurely jurisdictional.239 The Court proceeded to consider whether itwas appropriate for the federal courts to create a cause of action as a

232 This may explain the extraordinary degree of deference the courts accorded toexecutive branch suggestions of immunity for foreign states before the enactment ofthe FSIA. See infra notes 538–540 and accompanying text. R233 U.S. CONST. art. I, § 8, cl. 11.234 See Al-Bihani v. Obama, 619 F.3d 1, 19 (D.C. Cir. 2010) (Kavanaugh, J., concur-

ring in the denial of rehearing en banc) (“Sosa thus confirmed that international-lawprinciples are not automatically part of domestic U.S. law and that those principlescan enter into domestic U.S. law only through an affirmative act of the politicalbranches.”).235 See Ralph G. Steinhardt, Laying One Bankrupt Critique to Rest: Sosa v. Alvarez-

Machain and the Future of International Human Rights Litigation in U.S. Courts, 57 VAND.L. REV. 2241, 2254 (2004).236 See Young, supra note 34, at 28 (footnote omitted). R237 Sosa v. Alvarez-Machain, 542 U.S. 692, 712, 724 (2004).238 Id. at 713 & n.10.239 Id. at 714.



matter of federal common law.240 It concluded that it was appropriateto do so if certain conditions were met. In particular, the Court heldthat such a cause of action could properly extend only to norms ofcustomary international law that satisfy a heightened standard of clar-ity and breadth of acceptance.241 Bradley and Goldsmith, in theirpost-Sosa article written with Professor David Moore, stress that theselimits, and the Court’s generally cautious approach to this issue, areinconsistent with the modern claim that all customary internationallaw has the status of federal common law.242

But, as discussed above, the question whether customary interna-tional law has the status of federal law is distinct from the questionwhether someone injured by a violation of such law has a cause ofaction for damages against the individual who caused the injury. Evenif international law makes the defendant’s conduct unlawful, it doesnot itself typically establish a right of the victim to collect damagesfrom the individual who acted unlawfully.243 That the Court limitedthe cause of action for damages to norms having certain characteris-tics does not establish that only norms having these characteristicshave the force of federal law. In the absence of a cause of action,norms of customary international law could have the force of federallaw and be enforceable if invoked defensively or pursuant to a right ofaction created by another law, such as § 1983.244 It is not uncommonfor the Court to deny the existence of a federal right of action fordamages under federal statutes,245 yet no one claims that such statuteslack the force of federal law.

Bradley, Goldsmith, and Moore briefly consider and reject theargument that Sosa is inapposite to their critique of the modern posi-tion because the question before the Court was the existence of a pri-vate right of action for damages.246 They say that the opinion cannotfairly be read as turning on the fact that the issue was the existence of

240 Id. at 725–28.241 Id. at 724–25 (recognizing a federal right of action for violation of customary

international law norms that are “accepted by the civilized world and defined with aspecificity comparable to the features of the 18th-century paradigms” that Blackstonehad described, viz., “violation of safe conducts, infringement of the rights of ambassa-dors, and piracy”).242 Bradley, Goldsmith & Moore, supra note 33, at 902–07. R243 See supra notes 39–40 and accompanying text. R244 42 U.S.C § 1983 (2006).245 See, e.g., Karahalios v. Nat’l Fed’n of Fed. Emps., Local 1263, 489 U.S. 527,

536–37 (1989); Touche Ross & Co. v. Redington, 442 U.S. 560, 568, 578 (1979).246 Their first response is that this objection is unavailable to those who believe the

Court embraced the modern position. Bradley, Goldsmith & Moore, supra note 33, at R909. Presumably, it is still open to the rest of us.



a private cause of action because the Court extensively discussed post-Erie federal common law and cited the “limits on implied rights ofaction as only one of many reasons for judicial caution in allowingclaims under” § 1350.247 It is true that the Court cited five reasons tobe cautious in creating a federal right of action for violations of cus-tomary international law, but the fact that the issue before it waswhether “to create a private right of action” was prominent amongthem.248 The Court wrote:

The creation of a private right of action raises issues beyond themere consideration whether underlying primary conduct should beallowed or not, entailing, for example, a decision to permit enforce-ment without the check imposed by prosecutorial discretion.Accordingly, even when Congress has made it clear by statute that arule applies to purely domestic conduct, we are reluctant to inferintent to provide a private cause of action where the statute doesnot supply one expressly.249

And, although it is true that two of the other reasons for caution givenby the Court related to “post-Erie federal common law,”250 the propri-ety of implying a private right of action is commonly treated as anaspect of “post-Erie federal common law.”251

Furthermore, the Court’s final two reasons for caution, the onlytwo that relate specifically to the “foreign affairs” nature of the issue,apply only to a sort of case that falls outside the allocation-of-powersrationale for according customary international law the force of pre-emptive federal law252—a type of case as to which the Court in Sabba-tino had already articulated limits similar to those adopted in Sosa.253

The Court’s fourth reason for caution was based on the idea that[i]t is one thing for American courts to enforce constitutional limitson our own State and Federal Governments’ power, but quiteanother to consider suits under rules that would go so far as toclaim a limit on the power of foreign governments over their own citi-zens, and to hold that a foreign government or its agent has trans-gressed those limits.254

247 Id.248 Sosa v. Alvarez-Machain, 542 U.S. 692, 727 (2004).249 Id.250 See id. at 725–26.251 For example, the issue is covered in the “Federal Common Law” section of

Hart and Wechsler’s casebook on federal courts and the federal system. See HART &WECHSLER, supra note 55, at 690–742. R

252 See Sosa, 542 U.S. at 727–28.253 See Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398, 428 (1964).254 Sosa, 542 U.S. at 727 (citing Sabbatino, 376 U.S. at 431–32) (emphasis added).



The Court obviously had in mind the Filartiga-type case involving analien challenging the act of a foreign official, a sort of case fallingoutside the core rationale for treating customary international law aspreemptive federal law. It was for the same reason that the Court inSabbatino excluded disputed or controversial norms of customaryinternational law from the scope of those having the force of preemp-tive federal law in the context of suits challenging the conduct of for-eign states within their own territory.255 This limitation dovetails withthe Court’s fifth and final reason for caution: “We have no congres-sional mandate to seek out and define new and debatable violations ofthe law of nations . . . .”256 The Court in Sosa accordingly placed alimit on the judicially created federal right of action for damages forviolation of customary international law very similar to the limit thatthe Court had imposed in the parallel context in Sabbatino—a limitturning on the clarity and breadth of acceptance of the claimednorm.257 In sum, the Court’s reasons for caution are compatible withthe modern position not just because the question before it was thecreation of a private remedy not conferred by customary internationallaw ex proprio vigore, but also because some of the Court’s concerns areimplicated only in contexts that are not the central concern of themodern position.258

Although the Court in Sosa was not squarely presented with thequestion of customary international law’s status as federal or State law,its opinion tends to support the modern position. First, in explainingthat the Court’s “post-Erie understanding has identified limitedenclaves in which federal courts may derive some substantive law in acommon law way,” the Court recognized that, “[f]or two centuries wehave affirmed that the domestic law of the United States recognizes

255 See Sabbatino, 376 U.S. at 428.256 Sosa, 542 U.S. at 728 (emphasis added).257 See id. at 724–25; Sabbatino, 376 U.S. at 428; see also William S. Dodge, Customary

International Law and the Question of Legitimacy, 120 HARV. L. REV. F. 19, 26 (2007),http://www.harvardlawreview.org/media/pdf/dodge.pdf.258 The Court’s concern with imposing debatable rules on foreign state actors

should not have led it to restrict the right of action against Sosa, since Sosa was actingas an agent of the United States when he performed the acts on which Alvarez’s suitwas based. This latter fact may suggest that the Sosa limitations are not limited todamage actions alleging violations of international law attributable to foreign officials,as in the Filartiga-type case. On the other hand, the Court nowhere asserted thatSosa’s actions were attributable to the United States, but cf. Sosa, 542 U.S. at 737 (“Andall of this assumes that Alvarez could establish that Sosa was acting on behalf of agovernment when he made the arrest, for otherwise he would need a rule broaderstill.”), and some of the Court’s reasons for restricting the right of action are impli-cated only in the Filartiga-type case.



the law of nations.”259 In addition, it noted that the Court in Sabbatinohad “endorsed the reasoning” of Philip Jessup, “who had argued thatErie should not preclude the continued application of internationallaw in federal courts.”260

Moreover, Sosa’s willingness to create a federal common law rightof action for damages for violation of some norms of customary inter-national law itself has implications for the debate about the status ofcustomary international law, and those implications favor the modernposition. Given how strict the Court has been of late in recognizingnew federal rights of action, what is most striking about Sosa is its rela-tive receptivity to recognizing one in this context. In cases thataddress the existence of a federal right of action for damages for viola-tion of a federal statute that does not explicitly provide a right ofaction, the Court’s position is flatly that it will not create one.261

Rather, it will recognize only rights of action created by Congress.The Court’s holding in Sosa that § 1350 was “strictly jurisdic-tional”262—as well as its acknowledgement that the question before itwas whether to “create” a private right of action—means that theCourt did not understand the right of action for violations of custom-ary international law to have been created by Congress.263 Yet theCourt found it proper to create a right of action for violation of cer-tain norms of customary international law, noting that the interna-tional context made “the absence of congressional action addressingprivate rights of action . . . more equivocal than its failure to providesuch a right when it creates a statute.”264

In the absence of a federal right of action, persons injured byviolations of federal statutes are relegated to remedies that State lawmay provide. Thus, in Merrell Dow Pharmaceuticals Inc. v. Thompson,265

the Court found that an individual injured by a company’s violation of

259 Id. at 729 (citing Sabbatino, 376 U.S. at 423, The Paquete Habana, 175 U.S. 677,700 (1900), The Nereide, 13 U.S. (9 Cranch) 388, 423 (1815), and Tex. Indus., Inc. v.Radcliffe Materials, Inc., 451 U.S. 630, 641 (1981)). Since the Court was citing thesecases as supporting the propriety of recognizing a new federal common law cause ofaction, this statement suggests that it understood the cited cases to be treating cus-tomary international law as federal law, or at least as more like modern-day federal lawthan modern-day State law.260 Id. at 730 n.18.261 See Alexander v. Sandoval, 532 U.S. 275, 286–87 (2001) (“Without [Congres-

sional authorization], a cause of action does not exist and courts may not create one,no matter . . . how compatible with the statute.”).262 Sosa, 542 U.S. at 713.263 See Dodge, supra note 257, at 20–21. R264 Sosa, 542 U.S. at 727.265 478 U.S. 804 (1986).



standards set forth in the federal Food, Drug, and Cosmetics Act(FDCA), could sue only on a State law negligence per se theory.266 InSosa, the Court could have held that the right of action for damagesfor violation of customary international law came from State law. If ithad understood customary international law to be a State law matterafter Erie, it presumably would have taken this course. The fact that itdid not even consider the possibility of relegating persons com-plaining of international law violations to their State law remedies sug-gests that the Court was not inclined to regard customaryinternational law as a State law matter. The fact that the case involvedinternational law made the Court more willing—not less—to recognizea federal right of action under customary international law thanunder federal statutes, presumably because it considered it less appro-priate to leave the matter to State law for the reasons noted by Profes-sor Jessup and endorsed by the Court in Sabbatino.267

Finally, it is not the case that Sosa “confirmed that international-law principles are not automatically part of domestic U.S. law and thatthose principles can enter into domestic U.S. law only through anaffirmative act of the political branches,”268 or that Sosa’s recognitionof a right of action turned on “its reading of the specific intent ofCongress”269 or on “Congress’s intent in enacting” § 1350.270 TheCourt’s holding that § 1350 is purely jurisdictional contradicts anyclaim that Congress, in enacting that provision, intended to transformany norm of customary international law into “domestic U.S. law” orto create a right of action for damages for violation of such norms.Rather, Congress merely intended to allow independently existingcauses of action to be maintained in federal court. The First Con-gress’s belief that such causes of action existed was based on its under-standing of the nature and status of customary international lawwithin our legal system. Since the members of the Congress thatenacted § 1350 were also Founders, it is fair to say that the Sosa Courtbased its willingness to recognize a federal right of action for damagesfor certain violations of customary international law on the Founding

266 Id. at 807, 817; see also Blue Chip Stamps v. Manor Drug Stores, 421 U.S. 723,759–60 (1975) (Powell, J., concurring).267 Cf. Sosa, 542 U.S. at 730 n.18 (citing Banco Nacional de Cuba v. Sabbatino, 376

U.S. 398, 421–23, 425 (1964) (citing Jessup, supra note 185) (arguing that “Erie Rshould not preclude the continued application of international law in federalcourts”)).268 See Al-Bihani v. Obama, 619 F.3d 1, 19 (D.C. Cir. 2010) (Kavanaugh, J., concur-

ring in the denial of rehearing en banc).269 Young, supra note 34, at 29. R

270 Bradley, Goldsmith & Moore, supra note 33, at 896. R



generation’s understanding of the nature and status of customaryinternational law, not on any “affirmative act” of Congress. In otherwords, the Court in Sosa was not giving effect to the intent of Con-gress, but rather translating into post-Erie terms the Founders’ under-standing of the nature of the law of nations as it interacted with the“ambient law of the era.”271 That is also a fair description of whatdefenders of the modern position seek to do.272

* * * *

In sum, the constitutional structure strongly supports denying theStates the power to place the nation in breach of its obligations undercustomary international law. This structural case was well understoodby the Founders, who placed the responsibility for compliance withinternational law squarely in the hands of the federal government.The Founders gave Congress the authority to legislate regarding thelaw of nations, but they also contemplated that the courts would applysuch law in appropriate cases even in the absence of legislative incor-poration. They regarded the state-to-state portion of the law ofnations—the portion that corresponds to modern-day customaryinternational law—as either preemptive of State law or as having thestatus of general common law. Some Founders, in turn, understoodthe general common law to be federal law, while others regarded it asa category of law distinct from both federal and State law.

The pre-Erie cases reflect the same range of views regarding thestatus of the state-to-state branch of the law of nations. Some judgesclearly understood this law to be preemptive of State law for structuralconstitutional reasons. While the cases are equivocal as to whetherthe State courts’ decisions regarding such law were reviewable by theSupreme Court, even the cases calling such review into questionaffirm that such law was nevertheless binding on the States. The pre-

271 Sosa, 542 U.S. at 714.272 If § 1350 is purely jurisdictional, then it is also difficult to explain how the First

Congress believed the statute fell within the bounds of Article III unless it eitherunderstood the law of nations to be federal or embraced some theory of protectivejurisdiction. Since I accept protective jurisdiction, I do not read Sosa as necessarilyindicating that the First Congress regarded the law of nations as federal. Revisionistswho reject protective jurisdiction, see supra note 54 and accompanying text, can argue Rthat the First Congress overlooked Article III, but such arguments are disfavored. Butcf. Mossman v. Higginson, 4 U.S. (4 Dall.) 12, 14 (1800) (construing provision of FirstJudiciary Act conferring jurisdiction over suits “where an alien is a party” as requiringone of the parties to be a citizen of a State); Bellia & Clark, supra note 41, at 451 R(arguing that the Framers of § 1350 understood it to apply only to suits by aliensagainst U.S. citizens).



Erie doctrinal landscape thus falls far short of supporting the revision-ist claim that all of customary international law during this period wasregarded as nothing more than general common law. In any event,the courts’ pre-Erie approach to the general common law differed dra-matically from their post-Erie approach to the common law. Giventhat the contending pre-Erie positions were that the state-to-state por-tion of the law of nations either had the force of preemptive federallaw or had an intermediate status closer to modern-day federal lawthan to modern-day State law, the pre-Erie cases fail to support therevisionists’ conclusion that customary international law should nowbe regarded as having a State law status that was never in play.

After Erie, the Supreme Court in Sabbatino endorsed Philip Jes-sup’s view that Erie did not require that customary international law betreated as State law, noting that “rules of international law should notbe left to divergent and perhaps parochial [S]tate interpretations.”273

Even Judge Hand’s opinion in Bergman v. De Sieyes,274 often cited byrevisionists in support of a State law status for such law, acknowledgedthat a State’s violation of well-established rules of international lawmight present a federal question. The recent Samantar decisionshows that no one today embraces the view that revisionists mistakenlyattribute to Judge Hand.

To be sure, not all of customary international law has the force ofpreemptive federal law in all contexts. Sabbatino holds that somenorms of customary international law may not be applied to challengecertain acts of foreign states within their own territory, and The PaqueteHabana recognizes that some customary international law is subject to“controlling executive acts.” Adherents of the modern position donot dispute these propositions, which are consistent with the struc-tural constitutional case for regarding customary international law asbinding on State actors.

Finally, revisionists are wrong to claim that Sosa rejects the mod-ern position. To the contrary, the Court’s creation of a federal com-mon law cause of action for damages for violation of certain norms ofcustomary international law is more supportive of the modern posi-tion than of the revisionist position. The Court recognized that Sab-batino had endorsed Jessup’s view regarding Erie’s effect on the statusof customary international law, and, in concluding that it was appro-priate to create a federal common law right of action for damages forviolation of certain norms of customary international law, it relied onthe fact that, “[f]or two centuries we have affirmed that the domestic

273 Sabbatino, 376 U.S. at 425 (footnote omitted).274 170 F.2d 360 (2d Cir. 1948).



law of the United States recognizes the law of nations.”275 Thus,although the Court did not definitively adopt the modern position inSosa, it came close to doing so.

II. THE INTERMEDIATE THEORIES

Revisionists and defenders of the modern position agree that,before Erie, customary international law was regarded as (at least) gen-eral common law.276 Federal courts reached their own conclusionsabout the content of the general common law, but State courts werenot bound by federal interpretations.277 Revisionists and defenders ofthe modern position likewise agree that, after Erie, the general com-mon law no longer exists.278 If customary international law now hasthe status of domestic law in this country at all, it has to be eitherfederal law or State law. If customary international law is now federallaw, then this law preempts any inconsistent State law, State courts arebound by the federal courts’ views regarding the content of this law,and the Supreme Court may review State court decisions regardingsuch law. If customary international law is now purely a matter ofState law, then it does not preempt State law, and the State courts’views about its content would (apparently) bind the federal courts,including the Supreme Court. The Third Restatement takes the posi-tion that customary international law is generally federal law, with theeffects just noted;279 the revisionists maintain that, unless transformedinto federal law through political branch lawmaking, customary inter-national law is, at best, State law.280 But both agree that, after Erie,these are the only two possibilities.

More recently, scholars have argued that customary internationallaw can have an intermediate status between State and federal law.Professor Ramsey has proposed that customary international law be

275 Sosa, 542 U.S. at 729.276 See, e.g., RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW pt. 1, ch. 2 intro-

ductory note (1987) (“During the reign of Swift v. Tyson, 41 U.S. (16 Pet.) 1, 10 L.Ed.865 (1842), State and federal courts respectively determined international law forthemselves as they did common law, and questions of international law could bedetermined differently by the courts of various States and by the federal courts.”);Bradley & Goldsmith, Critique, supra note 9, at 822–23 (noting that customary interna- Rtional law was generally applied as part of the “general common law” before Erie);Koh, supra note 16, at 1830–31 (same); Young, supra note 19, at 374 & n.43, 393–94 & Rn.143 (collecting articles).277 See Bradley & Goldsmith, Critique, supra note 9, at 823. R278 Goodman & Jinks, supra note 16, at 472; Koh, supra note 16, at 1831; Stephens, R

supra note 16, at 397; Young, supra note 19, at 375. R279 See RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW § 111(1) & cmt. d.280 See supra note 15 and accompanying text. R



regarded as “nonpreemptive” federal law, meaning that it can beapplied as federal law whenever it does not conflict with State law.281

Professor Young has argued that customary international law canremain general law, to be applied in American courts in accordancewith choice-of-law rules, just as foreign law sometimes is.282 DeanAleinikoff has proposed that customary international law be regardedas federal law applicable in federal courts but not applicable in Statecourts.283

Bradley, Goldsmith, and Moore and Bellia and Clark do notdefend an intermediate status for customary international law, butthey argue that some subset of customary international law may betreated as federal law even though not incorporated into law by one ofthe forms of federal law listed in the Supremacy Clause (statutes andtreaties). Bellia and Clark set forth a structural constitutional basis foraccording a subset of customary international law preemptive force.284

Bradley, Goldsmith, and Moore argue that customary internationallaw may (sometimes) be incorporated as federal law by the Presidentacting alone and by the courts pursuant to jurisdictional statutes.285

On closer inspection, some of these intermediate positions turnout to be not so intermediate, and all but that of Bellia and Clarkencounter significant problems.

A. Ramsey’s Position

Professor Ramsey argues that customary international law shouldbe regarded as nonpreemptive federal law, a law that federal courtsmay apply, but only if there is no conflicting State law.286 (Even then,apparently the State courts need not apply this law.287) Ramsey arguesthat this position is supported by the text of the Constitution as under-stood by the Founders.288 He also argues that his proposal would rep-licate the effect that customary international law possessed as pre-Eriegeneral law.289 On closer analysis, however, nonpreemptive federallaw appears to be another name for State law. Moreover, Ramsey’stheory would produce acute anomalies, would be potentially very diffi-

281 See RAMSEY, supra note 17, at 350–51; Ramsey, supra note 17, at 584; infra Part RII.A.282 See Young, supra note 19, at 370; infra Part II.B. R283 See infra Part II.C.284 See infra Part II.D.285 See Bradley, Goldsmith & Moore, supra note 33, at 922. R286 See Ramsey, supra note 17, at 577, 584. R287 See RAMSEY, supra note 17, at 350. R288 See id. at 348–55.289 See Ramsey, supra note 17, at 558. R



cult to apply, and would render customary international law inapplica-ble where it is needed most. Ramsey’s textual and historical supportfor this highly problematic approach is more equivocal than heclaims.

1. Nonpreemptive Federal Law as State Law

To modern ears, the concept of nonpreemptive federal lawsounds like a contradiction in terms. We are accustomed to thinkingof federal law as by its nature preemptive of conflicting State law.With respect to customary international law, Ramsey reverses the rela-tionship between State and federal law, stipulating that this federallaw applies only where there is no conflict with State law.290 In theevent of a conflict between customary international law and State law,the courts must apply State law.291

Under this theory, the space for application of customary interna-tional law depends on how one determines whether a conflict exists.For example, if State law recognizes a cause of action but does notaffirmatively recognize an immunity for a sitting head of state, is therea conflict with a norm of customary international law entitling thehead of state to immunity? The answer would appear to depend onwhether the State’s law purports to be comprehensive, in the sensethat it permits all that it does not prohibit. If State law is comprehen-sive in this sense, then there would be a conflict between a State’sprovision of a cause of action and a norm of customary internationallaw providing immunity from such a cause of action, even if State lawdid not specifically deny such an immunity. Presumably, a federalcourt would answer the comprehensiveness question by reference toState law. If State law tells us that State law is comprehensive, thencustomary international law would not be applied at all unless theState affirmatively incorporated it. If State law tells us that State law isnot comprehensive with respect to customary international law, thenState law effectively incorporates customary international law whole-sale, subject to being overridden at retail by statute or perhaps byState judicial decision. In either case, customary international lawwould be applicable solely as State law.

Most likely, State law will be silent as to whether it is comprehen-sive. In such circumstances, the federal court could perhaps deter-mine that it is appropriate to apply customary international law to theextent that it does not directly conflict with State law. But thisapproach could easily be understood as the application of interna-

290 See id. at 584.291 See id.



tional law as State law. Indeed, this would appear to be a perfectlyreasonable interpretation of State law that is silent regarding interna-tional law. Since customary international law was regarded as part ofthe law of England, it is proper to assume that the States received suchlaw into their common law upon independence, or upon becoming aState, subject to legislative override.292 In short, if the federal courtsanswer the comprehensiveness question by reference to State law, itwould seem to follow that customary international law would be appli-cable as a matter of State law, not federal law.

On the other hand, for the federal courts to answer the compre-hensiveness question by reference to federal law would be to give cus-tomary international law the force of preemptive federal law, runningafoul of Ramsey’s insistence that such law is nonpreemptive. Ramseysays that, when State law does not “explicitly reject” the immunitydefense, “the federal courts [could] apply a version of the CharmingBetsy rule, interpreting State law not to violate international law ifanother construction of the law is possible.”293 But, if Ramsey wouldapply such a rule even when State law purports to be comprehensiveand does not provide for such a rule, then he would be allowing cus-tomary international law to preempt State law. Because he insists thatcustomary international law is nonpreemptive, presumably he wouldnot allow the presumption to be applied in such circumstances. If, onthe other hand, he would allow federal courts to apply the CharmingBetsy rule only when State law does not purport to be comprehensive,then, again, the applicability of customary law would be fully explica-ble on the theory that the common law, as received by the State fromEngland, presumptively incorporates customary international law. Onthis theory, the federal courts would be applying customary interna-tional law as State law, not federal.

In any event, the States would be free to declare that their courtswill not construe State statutes or the common law to conform to cus-tomary international law, as the voters of Oklahoma recently havedone.294 Indeed, the States would presumably be free to declare thatthey will construe State law to violate customary international lawwhenever possible. Ramsey says that “the [Charming Betsy] presump-tion is not the same as preemption, so its use is not governed by Arti-cle VI.”295 But he elsewhere makes clear that State law prevails in the

292 See Young, supra note 19, at 480. R

293 See Ramsey, supra note 17, at 578 (referring to Murray v. Schooner Charming RBetsy, 6 U.S. (2 Cranch) 64, 118 (1804)).294 See supra note 1 and accompanying text. R

295 Ramsey, supra note 17, at 578. R



event of a conflict.296 If State law specifies that State law is not to beinterpreted to conform to international law, then there would appearto be a conflict between State law and customary international lawwhenever the Charming Betsy presumption would otherwise be doingany work. Thus, Ramsey’s claim that “federal courts might have thepower to interpret [S]tate law to avoid . . . conflicts [between State lawand customary international law], even while recognizing that in theevent of a conflict the [S]tate would prevail,”297 would appear to holdtrue only if the State does not itself provide for a different way toresolve such conflicts. And if State law does not provide for a differentpresumption, then the Charming Betsy presumption could be under-stood as a State law presumption. As such, it would be fully consistentwith the revisionist view.

Customary international law appears to be doing more work inone of Ramsey’s examples. Ramsey posits a claim under customaryinternational law “against a multinational corporation for complicityin human rights violations in connection with a project in Burma.”298

Ramsey might permit a federal court to resolve this dispute on thebasis of customary international law, but only if the relevant State con-flict-of-laws rules would not apply State law to this claim, but wouldinstead apply Burmese law, and Burmese law would not support recov-ery.299 The federal court could recognize the claim even if the Statecourts would not recognize a claim based on international law.300 It isnot clear why Ramsey would accept the applicability of customaryinternational law to this case. He tells us that he would not permit thefederal court to adjudicate the claim under international law if, underState choice-of-law rules, the State courts would apply State law.301

Presumably that is because he would regard the State choice-of-lawrule as effectively extending State law extraterritorially to this dispute.But doesn’t a State choice-of-law rule making Burmese law applicableto this dispute reflect the State’s determination that this dispute isproperly governed by Burmese law? If a State’s courts have deter-mined that the dispute is properly resolved under Burmese law, theState is not neutral with respect to the applicability of internationallaw; presumably, the State would regard it as proper to resolve thedispute according to international law only if Burmese law incorpo-rated international law. But Ramsey would permit the federal courts

296 See id. at 585.297 Id. at 579.298 Id.299 See id.300 Id.301 See id. at 579–80.



to apply customary international law even if Burmese law does notincorporate it.302

Ramsey claims that the federal court would not be displacingState choice-of-law rules because “state conflicts rules do not addressthe status of international law; these rules only decide which munici-pal legal rules to apply.”303 But this is true only with respect to hori-zontal choice-of-law rules. Whether to apply a rule of municipal lawthat conflicts with international law is a vertical choice-of-law question.If a State’s courts would not apply international law to a case, that ispresumably because the State has determined that the dispute is prop-erly resolved according to municipal law even if it conflicts with inter-national law. Ramsey’s reverse-supremacy approach to customaryinternational law would appear to require a federal court to do thesame. To do otherwise would be to replace the State’s vertical choice-of-law rule with a federal one.

Ramsey also argues that a federal court that applied internationallaw would not be displacing the State’s refusal to recognize the inter-national law claim because such a refusal is “just a decision by the[S]tate that its courts are not a forum for international law claims.”304

But a decision not to provide a forum for the adjudication of the dis-pute would usually be framed as a dismissal for lack of jurisdiction oron the ground of forum non conveniens, neither of which would haveres judicata effect.305 A court’s resolution of the dispute according toBurmese law would have res judicata effect, and would thus bar a sub-sequent claim under international law. If so, a State choice-of-law rulewould appear to bear on substance and presumably have to be fol-lowed by federal courts.306

At any rate, the States would be free to characterize their choice-of-law rules as substantive. Thus, at best, federal courts would be freeto apply international law only if the States characterized their choice-of-law rules as nonsubstantive denials of a forum. But, again, theStates would retain ultimate control over the applicability of interna-tional law in federal courts, as Ramsey implicitly recognizes when heconcedes that the federal courts would not be free to apply interna-tional law if the States made it clear that their refusal to recognize an

302 Id. at 579.303 Id.304 Id.305 See Mizokami Bros. of Ariz. v. Mobay Chem. Corp., 660 F.2d 712, 716–17 (8th

Cir. 1981); cf. Parsons v. Chesapeake & Ohio Ry. Co., 375 U.S. 71, 72–73 (1963) (percuriam) (holding that a State court dismissal on forum non conveniens grounds doesnot preclude federal court’s consideration of § 1404(a) transfer).306 See Klaxon Co. v. Stentor Electric Mfg. Co., 313 U.S. 487, 496–97 (1941).



international law claim was based on their determination that some orall corporations were, as a substantive matter, not subject to interna-tional law.307

Ramsey’s approach would also produce significant problems ofapplication. First, customary international law would presumably beinapplicable if it conflicted with the law of any of the fifty States. (Ifcustomary international law is nonpreemptive, then presumably itcannot preempt the law of any of the fifty States.) Under Ramsey’sapproach, therefore, the federal courts would be required to canvassthe laws of all States whose laws could constitutionally be applied tothe dispute to determine whether they would regard their law as appli-cable and whether their law conflicts with customary international law.Since, as Ramsey recognizes, the constitutional limits on the applica-tion of State law to foreign disputes are quite modest,308 the numberof States that may constitutionally apply their law to any given disputecould be large. The analysis would appear to require the resolution ofcomplex choice-of-law issues under various States’ laws, as well asissues of substantive law.

Second, if Ramsey would indeed permit a federal court to enter-tain an international law claim when the State courts would not (as heclaims he would), his approach would replicate the forum shoppingproblem that Erie sought to eliminate. If nonpreemptive federal lawwould not support “arising under” jurisdiction (a question that Ram-sey leaves open309), then his approach would result in federal courtspossibly reaching a different result than State courts depending onthe “accident of diversity.”310 Ramsey acknowledges the problem, butmaintains that it is a necessary consequence of the rule adopted by theConstitution (a claim I consider below).311

Ramsey perceives, and accepts, a second problem with his propo-sal: that it would produce a lack of uniformity between State and fed-eral courts in the interpretation of customary international law.312

Ramsey argues that nonuniformity should not be regarded as a prob-lem because “[n]on-uniformity was the rule for at least the 150-oddyears prior to Erie” and it did not create significant foreign relations

307 See Ramsey, supra note 17, at 579 n.91. R

308 See id. at 573.309 See id. at 583 n.105.310 Klaxon, 313 U.S. at 496.311 See Ramsey, supra note 17, at 583. R

312 Id. at 582 (noting that the nonuniformity problem arises “if the federal courtsand State courts reach independent conclusions as to the substantive content of inter-national law”).



problems then.313 For reasons explained in Part IV, divergent Stateand federal interpretations of customary international law wouldindeed pose a significant foreign relations problem for the nation.On the other hand, it seems to me that Ramsey’s approach does notpresent the problem of nonuniformity between a federal court andthe courts of the State in which it sits. Ramsey’s acceptance of thisobjection suggests that he envisions that federal courts will not bebound by State court interpretations of customary international law.But, if customary international law were nonpreemptive, then it wouldseem to follow that federal courts will in fact have to follow State courtinterpretations of such law. Ramsey admits that federal courts cannotapply customary international law if the States have rejected it. It fol-lows that, if a State incorporates customary international law in partbut rejects it in part, then federal courts may enforce only the partthat the State has incorporated. It follows further that, if the State hasincorporated a mistaken interpretation of customary internationallaw, the federal courts are bound to enforce that mistaken interpreta-tion. The States could in theory decide to incorporate customaryinternational law as interpreted by the federal courts; this wouldamount to a delegation of the interpretation of customary interna-tional law by the State to the federal courts. In this situation, therewould be no lack of uniformity. Apart from such a delegation, theonly time the federal courts would be using their independent judg-ment about the content of customary international law would be whenState law is silent about its content. Since, by hypothesis, there wouldbe no State interpretation of international law in such contexts, therewould be no lack of uniformity. In sum, lack of uniformity as betweenState and federal courts is not a problem, but only because Ramsey’sapproach is, in application, indistinguishable from the revisionist viewthat customary international law is State law.

On the other hand, Ramsey’s approach would produce a differ-ent sort of disuniformity: the fifty different States would each reachtheir own conclusions regarding customary international law, thuspotentially producing fifty different interpretations of customaryinternational law. Indeed, the federal courts would themselves beapplying nonuniform interpretations of customary international law,since they would have to follow the nonuniform interpretations ofsuch law by the fifty State courts. And federal and State courts alikewould confront potentially complex choice-of-law problems if differ-ent States adopted different interpretations of international law. TheSupreme Court in Sabbatino expressed concern about such

313 Id.



nonuniformity when it wrote that “rules of international law shouldnot be left to divergent and perhaps parochial [S]tateinterpretations.”314

It is not true that we managed to live with such disuniformity for150 years. As discussed in Part I, during those years, customary inter-national law was regarded as either binding and preemptive federallaw or as general common law. To the extent it was regarded as thelatter, federal and State courts alike understood their task to be todiscover and apply a general law whose content did not depend ontheir decisions, a regime that resulted in a remarkable degree of uni-formity.315 By contrast, if Ramsey’s approach were adopted, the Statecourts would be accorded the power to incorporate customary inter-national law—and to require or preclude the federal courts fromapplying it—in whole, in part, or not at all. The State courts wouldalso be recognized to have the power to modify or supplement suchlaw, and again the federal courts would be required to follow. Con-flicting interpretations—and conflicting degrees of incorporation—are much more likely to occur if Ramsey’s approach were adoptedthan under the pre-Erie general law approach.

Finally, and most importantly, Ramsey’s approach would disallowfederal court enforcement of customary international law in preciselythe type of case that concerned the Founders the most. As many haverecounted, the Founders were most concerned about violations ofinternational law by the States.316 The problems caused by State viola-tions of international law demonstrated to the Founders that responsi-bility for compliance with international law had to be assigned to thefederal government.317 Ramsey recognizes the structural problemwith leaving questions of international law to the States, but he arguesthat the Founders addressed it, with respect to customary interna-tional law, by giving Congress the power to require States to complywith such law.318 Unless Congress gives preemptive effect to custom-ary international law, he argues, the States retain the power to violateit.319 I discuss Ramsey’s basis for these conclusions below. The pre-

314 Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398, 425 (1964). The Courtattributed this concern to Professor Philip Jessup, see id., but Jessup’s article does notexpress a concern about nonuniformity, see Jessup, supra note 185, at 743. It is thus Rlikely that it was the Court itself that perceived that divergent State court internationallaw would be a problem.315 See supra note 178 and accompanying text. R316 See, e.g., Bradley, supra note 86, at 144. R317 Id. at 146.318 See Ramsey, supra note 17, at 582. R319 RAMSEY, supra note 17, at 350. R



sent point is that Ramsey’s argument that customary international lawis nonpreemptive federal law is unresponsive to the Founders’ mainconcerns. Under Ramsey’s approach, the federal courts wouldremain powerless in the face of State action contravening such law.They would be able to enforce such law only if the States have beensilent, and they would be able to resolve an international law claimdifferently from the State courts only in the unusual situation of adispute arising from conduct that took place entirely abroad—a typeof case that was hardly a central concern of the Founders. It seemshighly unlikely that these very unusual situations were what the Foun-ders had in mind when they wrote, for example:

Under the national government . . . the laws of nations . . . willalways be expounded in one sense and executed in the same man-ner—whereas adjudications on the same points and questions inthirteen States . . . will not always accord or be consistent . . . . Thewisdom of the convention in committing such questions to the juris-diction and judgment of courts appointed by and responsible onlyto one national government cannot be too much commended.320

2. Ramsey’s Textual and Historical Support

Ramsey acknowledges that, under his approach, the applicabilityof international law remains within the States’ control.321 He insiststhat this necessarily follows from the constitutional text and the Fram-ers’ intent.322 But neither the text nor the Framers’ intent providessignificant support for his argument.

Ramsey’s principal textual argument is that the Constitution enu-merates the categories of preemptive federal law and customary inter-national law is not among them.323 This omission is particularlytelling, he argues, because the other principal form of internationallaw—treaties—is listed.324 In addition, he notes that the Constitutionis not silent about customary international law: it expressly gives Con-gress the power to “define and punish . . . Offences against the Law ofNations.”325 He argues that the Offenses Clause tells us that custom-ary international law can be preemptive federal law only if enacted

320 THE FEDERALIST NO. 3, at 37–38 (John Jay) (Clinton Rossiter ed., 1961).321 See Ramsey, supra note 17, at 582. R322 See id. at 585.323 See id. at 559–60.324 Id. at 560; see also U.S. CONST. art. VI, cl. 2 (“This Constitution, and the Laws of

the United States . . . and all Treaties made, or which shall be made, under the Author-ity of the United States, shall be the supreme Law of the Land . . . .” (emphasisadded)).325 U.S. CONST. art. I, § 8; see Ramsey, supra note 17, at 560. R



into federal law by statute (or treaty).326 In other words, customaryinternational law is non-self-executing federal law.

But there is another possible explanation for the inclusion oftreaties but not customary international law in the Supremacy Clause.The British rule with respect to treaties, which we would have inher-ited had the Constitution not established a different rule, was thattreaties do not have the force of domestic law until implemented bystatute.327 One purpose of the Supremacy Clause with respect to trea-ties was to reverse the British rule and give treaties the effect ofdomestic law once in force.328 With respect to customary interna-tional law, the British rule was the opposite: such law was regarded aspart of the law of the land without the need for legislative implemen-tation.329 If the Founders assumed that the British approach wouldapply unless changed, then there was no need to declare that custom-ary international law was to have the force of domestic law. Even with-out such a declaration, customary international law would be the lawof the land, just as it was in England.

Of course, a second effect of the Supremacy Clause with respectto treaties was to establish that they were preemptive of State law.Because customary international law is not listed, the text of theSupremacy Clause does not establish that such law is likewise preemp-tive of State law. Ramsey argues that such law cannot be understoodto have been included in Article VI’s reference to the “laws of theUnited States . . . made in Pursuance [of the Constitution],” sincecustomary international law is not made through the lawmaking pro-cedures set forth in the Constitution.330 But, as noted above, someFounders appear to have regarded the preemptive effect of (some of)the law of nations as an inference from the constitutional structure,and the Constitution is, of course, part of the supreme law of theland.331 Moreover, the text cannot help Ramsey if, as revisionistsargue, customary international law was understood by the Founders tobe general law.332 Pre-Erie general law had some characteristics ofcontemporary federal law (federal courts were not bound by Statecourt interpretations of the law) but lacked others (State courts could

326 See RAMSEY, supra note 17, at 348–50; Ramsey, supra note 17, at 560–61. R327 See Vazquez, Treaties as Law of the Land, supra note 86, at 614–15. R328 See generally id. at 616–19 (describing the purpose and original meaning of the

Supremacy Clause).329 See 4 WILLIAM BLACKSTONE, COMMENTARIES *67.330 See Ramsey, supra note 17, at 559–60 (quoting U.S. CONST. art. VI, cl. 2). R331 See supra text accompanying note 117; see also infra text accompanying note R

350. R332 See Young, supra note 19, at 374 & n.43. R



enforce their own interpretations of the law).333 If that intermediateoption is no longer available, then the status of customary interna-tional law, post-Erie, must change in one direction or the other. If thetext of the Supremacy Clause was understood to leave customary inter-national law with that intermediate status, then that text cannot helpus choose between the two other options.

Ramsey defends his proposal as the closest available approxima-tion of the intermediate status that customary international law pos-sessed before Erie, but his proposal would make customaryinternational almost indistinguishable from State law. His attempt torecapture customary international law’s pre-Erie status overlooks hisown insight that Erie “did not just overrule one line of cases, but over-threw an entire way of thinking about law.”334 As noted above, beforeErie, State and federal interpretations of the general common law con-verged because those courts believed themselves to be involved in thecommon enterprise of discovering and applying a law whose contentdid not depend on their decisions. With that way of thinking aboutthe law overthrown, relegating customary international law to the sta-tus of the present-day common law would produce far greater dis-uniformity than the Founders contemplated. In light of the Statecourts’ new conception of their role vis-a-vis the common law, custom-ary international law’s pre-Erie status would be more closely recap-tured if it were regarded as federal law.335

The Offenses Clause similarly fails to support Ramsey’s position.Revisionists read that clause to establish that the federal law status ofcustomary international law depends on legislative transformation.336

But, here again, an alternative reading is available. The OffensesClause can be understood instead to reflect the view that congres-sional action is necessary to make violations of customary interna-tional law a criminal offense. Although today it is accepted that theOffenses Clause empowers Congress to pass civil as well as criminallaws,337 it is clear from the clause’s use of the terms “punish” and“Offenses” that criminal laws were the central concern of theclause.338 Today, it is well accepted that the violation of treaties can-

333 See Bradley & Goldsmith, Critique, supra note 9, at 823. R334 Ramsey, supra note 17, at 585. R335 For elaboration of this point, see supra text accompanying note 178. R336 See Bradley & Goldsmith, Current Illegitimacy, supra note 2, at 325; Young, supra

note 19, at 393. R337 See J. Andrew Kent, Congress’s Under-Appreciated Power to Define and Punish

Offenses Against the Law of Nations, 85 TEX. L. REV. 843, 861–62 (2007).338 Cf. id. at 868 (noting that the executive branch until recently interpreted the

Offenses Clause to permit proscription of criminal conduct only).



not be made a criminal offense except pursuant to statute.339 A forti-ori, the same must be true of customary international law. Sincetreaties are concededly federal law even though a federal statute isrequired to make violations criminal, the Constitution’s enumerationof an express federal power to criminalize violations of customaryinternational law does not establish that customary international law isnon-self-executing in other respects.340

In addition to his textual analysis, Ramsey finds evidence of theFramers’ intent in Justice Chase’s opinion in Ware v. Hylton.341 Heargues that the following language from Chase’s opinion clearly estab-lishes that the Founders understood that customary international lawwould not preempt State law:

It is admitted, that Virginia could not confiscate private debts with-out a violation of the modern law of nations, yet if in fact, she has sodone, the [state] law is obligatory on all citizens of Virginia, and onher Courts of Justice; and, in my opinion, on all the Courts of theUnited States. If Virginia by such conduct violated the law ofnations, she was answerable to Great Britain, and such injury couldonly be redressed in the treaty of peace.342

In context, however, that language does not support the claim thatthe Founders believed that State law could preempt customary inter-national law. Justice Chase was discussing the petitioner’s argumentthat a Virginia statute confiscating the property of British citizens wasinvalid because it violated the law of nations.343 The relevant Virginialaw was enacted in 1777, before the Constitution was written, and evenbefore the Articles of Confederation were in force.344 Justice Chasemade it clear in his opinion that, in his view, Virginia at the time was asovereign nation.345 Thus, his statement that a statute of the Virginialegislature would prevail over customary international law stands onlyfor the proposition that a nation’s legislature may preclude the appli-cation of customary international law by that nation’s courts. The

339 See Carlos Manuel Vazquez, The Four Doctrines of Self-Executing Treaties, 89 AM. J.INT’L L. 695, 718 & n.108 (1995).340 It is true that, in the early Republic, it was believed by some that the common

law, unaided by a statute, could be the basis of a federal criminal prosecution. Still, itis likely that many Founders held the opposite view, a view that was later adopted bythe Supreme Court in United States v. Hudson & Goodwin, 11 U.S. (7 Cranch) 32(1812).341 3 U.S. (3 Dall.) 199 (1796).342 RAMSEY, supra note 17, at 352 (alteration in original) (citing Ware, 3 U.S. (3 R

Dall.) at 229 (opinion of Chase, J.)).343 See Ware, 3 U.S. (3 Dall.) at 222 (opinion of Chase, J.).344 See id. at 199–200 (statement of the case).345 See id. at 223–24 (opinion of Chase, J.).



equivalent proposition today would be that Congress has the powerunder our Constitution to preclude the application of customaryinternational law in our courts, a proposition that even the Restatementendorses.346 The structural argument for the preemptive force of cus-tomary international law relies on the place of the States in our fed-eral Union. Chase’s analysis in Ware does not contradict thatargument.347

Indeed, Chase’s opinion in Ware suggests that the Founderswould not have agreed that the States had the power to preempt cus-tomary international law under the Constitution as eventuallyadopted, or even under the Articles of Confederation. The petitionerin Ware argued alternatively that Virginia’s violation of the law ofnations was invalid as an interference with the national government’sexclusive power over war and peace.348 Chase rejected that argument,but only because he believed that, before the Articles of Confedera-

346 See RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW § 115(1)(a) (1987).Ramsey also relies on similar statements by Justice Iredell in his circuit court opinionin Ware, which was reversed by the Court on other grounds. See RAMSEY, supra note10, at 352–53 (citing Ware, 3 U.S. (3 Dall.) at 265–66 (opinion of Iredell, J.)). ButIredell, too, relied on the fact that Virginia’s law was passed before the Articles ofConfederation were in force: “[N]o one has suggested, that the act of October, 1777,was in any manner inconsistent with the Constitution of the [S]tate; and at that timethe articles of Confederation were not in force . . . .” Ware, 3 U.S. at 266 (opinion ofIredell, J.); cf. infra note 351. In addition, Ramsey relies on this statement from Jus- Rtice Cushing’s brief opinion: “A State may make what rules it pleases; and those rulesmust necessarily have place within itself. But here is a treaty, the supreme law, whichoverrules all State laws upon the subject, to all intents and purposes; and that makesthe difference.” RAMSEY, supra note 17, at 353 (quoting Ware, 3 U.S. (3 Dall.) at 282 R(opinion of Cushing, J.)). It is unclear how this statement supports Ramsey’s posi-tion. Clearly Cushing could not have meant that State laws are valid unless they vio-late a treaty; as Ramsey recognizes, State laws that violate the Constitution and federalstatutes are also invalid. In any event, in light of the reasoning of Chase and Iredell,the very general statement in Cushing’s opinion cannot be read to suggest that a Statestatute enacted after the adoption of the Constitution would be valid and binding even if itviolated customary international law.347 When Chase further indicated that the U.S. courts would similarly be pre-

cluded from invalidating Virginia’s act in reliance on customary international law, seeWare, 3 U.S. (3 Dall.) at 229 (opinion of Chase, J.), he presumably meant that the U.S.courts were required to resolve the validity of the Virginia confiscation pursuant tothe law applicable in Virginia at the time of the law’s enactment, and since neither theArticles of Confederation nor the Constitution was in force at the time, there was thenno higher domestic law to require the application of customary international law.Chase considered separately whether the laws or treaties that came into existenceafterwards retroactively required the invalidation of Virginia’s statute, and he con-cluded (along with all of the other Justices save Iredell) that they did. See id. at242–43.348 See Ware, 3 U.S. (3 Dall.) at 219–20 (argument of counsel for the petitioner).



tion came into force, Congress did not possess exclusive power overwar and peace.349 He recognized, however, that if the Constitutiondid allocate exclusive power over war and peace to the national gov-ernment, the States would have been implicitly prohibited from exer-cising “external sovereignty.”350 Ware thus reflects the view of at leastsome Founders that State discretion to violate international law isinconsistent with a federal structure.351 As noted above, the court inRutgers v. Waddington reached the same conclusion even with respectto the weaker federal system established by the Articles ofConfederation.352

Ramsey claims that his concept of nonpreemptive federal law rep-licates the pre-Erie concept of general law.353 It is true that most gen-eral law—the part of it that corresponds to what we now consider thecommon law—was thought to be alterable by State legislatures. But,as discussed above, it is far from clear that the part of general law thatcorresponds to today’s customary international law was also thoughtto be alterable by the States. Ramsey notes that there were no nine-teenth-century decisions striking down State statutes that conflictedwith customary international law,354 but he cites no cases upholdingsuch statutes either. The lack of cases either way suggests that theStates did not regard themselves as free to depart from that portion of

349 See id. at 231–33 (opinion of Chase, J.).350 See id. at 232. This is, of course, a precursor to the structural argument that

Professors Bellia and Clark develop in their article. See Bellia & Clark, supra note 21, Rat 5–7.351 In his opinion in Ware, Justice Iredell opined that if the Articles of Confedera-

tion had been in force in 1777:I think there is no colour for alledging [sic] any inconsistency [of the Vir-ginia law] with them, since Congress could have passed no act on this sub-ject, but if they had wished for an act, must have recommended to the StateLegislatures to pass it. And the very nature of a recommendation implies,that the party recommending cannot, but the party to whom the recommen-dation is made, can do the thing recommended.

Ware, 3 U.S. (3 Dall.) at 266 (opinion of Iredell, J.). If Iredell meant that the Virginialaw would have been valid under the regime of the Articles of Confederation even if ithad violated the law of nations, his dictum would be in tension with Rutgers, but itwould not imply that he would have considered such a law to be valid under theConstitution, since the Constitution does give the federal government the power topass a law on the subject and, indeed, all of the Justices in Ware found that the Consti-tution itself gave legal force to a treaty on the subject. See id. at 236 (opinion ofChase, J.); id. at 249 (opinion of Paterson, J.); id. at 259 (opinion of Iredell, J.); id. at281 (opinion of Wilson, J.); id. at 282 (opinion of Cushing, J.).352 See supra text accompanying note 117 (discussing Rutgers v. Waddington). R

353 See RAMSEY, supra note 17, at 585. R

354 See id. at 352.



the law of nations. This view may in turn have been based on theStates’ understanding that international law generally was binding onall, combined with a recognition that, if violations of international lawwere permitted at all by our Constitution, they would have to beauthorized by the federal government. As we have seen, the idea that,in a federal system, departures from international law must be a mat-ter for the federal government, not the States, even predated theConstitution.355

B. Young’s Position

Like Ramsey, Young proposes a solution that seeks to avoid theproblems of both the modern and revisionist positions by replicatingthe pre-Erie status of customary international law as general law.356

Unlike Ramsey’s proposal, Young’s is truly intermediate in that itwould not effectively relegate customary international law to State lawstatus. That is because Young would recognize that, in certain circum-stances, federal law may require the application of customary interna-tional law even if State law would not.357 Still, Young’s proposal tiltsheavily toward the revisionist view, as he makes it clear that the appli-cability of customary international law would ordinarily be deter-mined by State law.358 Moreover, the particular mechanism thatYoung proposes to determine when customary international law willbe applied in federal or State courts—the application of choice-of-lawrules359—is problematic in a number of respects.

355 See supra text accompanying note 117. R356 See Young, supra note 19, at 370. R357 See id. Indeed, Ramsey’s main criticism of Young’s proposal is that it would

effectively accord customary international law preemptive force in some circum-stances. See Ramsey, supra note 17, at 567–71. Young, for his part, criticizes Ramsey Rfor sometimes permitting Supreme Court displacement of State court applications ofcustomary international law. See Young, supra note 19, at 484. This criticism seems Rmisplaced. First, it depends on Young’s belief that Ramsey would regard customaryinternational law as federal law for purposes of arising under jurisdiction, see id. at483–84, whereas Ramsey expressly reserves judgment on that point, see Ramsey, supranote 17, at 583 n.105. Second, since Ramsey insists that customary international law is Rnonpreemptive, it would seem to follow that, even if the Supreme Court could exer-cise appellate jurisdiction over a case “arising under” that law, it may not review orreverse the State court’s interpretation of it. Indeed, although Young denies it, thelogic of his proposal would appear to require the conclusion that the Supreme Courtmay review State court applications of international law in certain circumstances. Seeinfra text accompanying note 448–49. In my view, Ramsey’s position is highly prob- Rlematic, and Young’s slightly less so, precisely insofar as they would deny the SupremeCourt authority to review State court interpretations of customary international law.358 See Young, supra note 19, at 508–09. R359 See id. at 468.



Section 1 considers Young’s criticisms of the modern position.Section 2 considers Young’s prescriptions.

1. Young’s Criticisms of the Modern Position

Young offers a lengthy critique of the modern position, which hedescribes as the view that “customary international law is always fed-eral law and always displaces [State] law, without consideration of thenature of the particular rule at issue or of the constellation of federal,[S]tate, international, and private interests that might be implicatedby it.”360 He finds “the revisionist critique of the modern position fun-damentally persuasive—at least in its negative project of demonstrat-ing that the wholesale incorporation of customary international law asfederal common law is inconsistent with constitutional principle.”361

He concludes that “the modern position offends constitutional normsof federalism, separation of powers, and democracy.”362

The critiques based on democracy, federalism, and separation ofpowers are overdrawn. Revisionists often criticize international lawfor its “democratic deficit.” One version of this critique stresses thatits norms often do not reflect the preferences of the majority of theworld’s people.363 Many countries that participate in its creation arenot democracies; the “dead hand” of older customary internationallegal norms binds states even after those norms have changed; andundemocratically selected elites, such as judges on international tribu-nals and law professors, have undue influence in the creation andinterpretation of customary international law.364 Young credits a dif-ferent version of the democracy critique: the makers of customaryinternational law do not fairly represent the people of the United States.Thus, quoting Professor Trimble, he notes that “at least some of thepotential lawmakers, such as foreign governments, are neither repre-sentative of the American political community nor responsive to it.”365

360 Id. at 437; see also id. at 444, 463, 509 (describing similarly the modern posi-tion). As discussed in subpart I.B, few proponents of the modern position take suchan uncompromising position.361 Id. at 462.362 Id.363 See J. Patrick Kelly, The Twilight of Customary International Law, 40 VA. J. INT’L L.

449, 518–23 (2000); John O. McGinnis & Ilya Somin, Democracy and InternationalHuman Rights Law, 84 NOTRE DAME L. REV. 1739, 1771–73 (2009).364 See McGinnis & Somin, supra note 363, at 1764–68. R365 Young, supra note 19, at 398–99 (quoting Phillip R. Trimble, A Revisionist View R

of Customary International Law, 33 UCLA L. REV. 665, 721 (1986)); see also Bradley &Goldsmith, Critique, supra note 2, at 857, 868 (arguing the modern position conflictswith basic principles of American democracy).



Young is not comforted by the “internationalist” response that the fed-eral political branches at least participate in the international lawmak-ing process.366

In my view, both versions of the democracy critique are mis-guided. As Young recognizes, compliance with international law wasan “important desideratum” for the Framers,367 who “clearly intendedthat customary international law should have application in ourcourts.”368 Yet, at the time of the Founding, far fewer of the world’snations were democracies than today—and, to the extent that the lawof nations was thought to be “made” at all, it was made at a muchearlier time, when even fewer countries were democracies. Moreover,in the Founders’ view, much of the law of nations was not “made” atall, but instead had its basis in natural law or right reason.369 Theywished to see international law complied with, not because it satisfiedtheir conceptions of democracy, but because they feared the conse-quences for the nation if such law were violated.370 As discussedabove, this remains the most powerful structural argument for treat-ing customary international law as preemptive federal law. The realitythat violations of international law risk bad consequences for thenation in the form of adverse reactions by other nations is a fact aboutthe world that must be taken into account in some fashion. (Moreo-ver, it bears recalling, in this connection, that adherents of the mod-ern position do not claim that customary international law’s status asfederal law is immune from alteration by democratic means. Thedebate concerns the default rule: does it take an act of federal politi-cal branches to prohibit States from violating such law, or to permitthem to do so?)

Young ultimately concludes that the democracy critique addsnothing to the federalism and separation-of-powers critiques.371 In

366 See Young, supra note 19, at 399. R367 Id. at 469 n.531.368 Id. at 390.369 See RAMSEY, supra note 17, 344–46; Saikrishna Prakash, The Constitutional Status R

of Customary International Law, 30 HARV. J.L. & PUB. POL’Y 65, 70 (2006). For thisreason, Young is on weak ground when he distinguishes the “old” law of nations fromthe “new” international law on the ground that the former was “empirically-based,”where as the latter is not. See Young, supra note 19, at 390–91. The “new” interna- Rtional law may or may not have an empirical basis, but it is clear that much of the“old” international law did not. In any event, this distinction is orthogonal to theFounders’ concerns. As discussed in the text, the Founders wished to see norms ofinternational law complied with, regardless of their provenance, because they wereconcerned about the consequences of violations of such law.370 See Stephens, supra note 16, at 400–02. R371 See Young, supra note 19, at 400. R



my view, the democracy critique is actually in tension with the federal-ism and separation-of-powers critiques. The latter critiques reducethemselves to the objection that customary international law is notmade in the manner contemplated by the Constitution for makingpreemptive federal law. Young relies heavily on the scholarship ofBradford Clark, who has argued forcefully that the list of categories ofsupreme federal law set forth in the Supremacy Clause is exclusive.372

As discussed above, Clark has argued that the process for makingthese three forms of federal law—such as the bicameralism and pre-sentment requirements for making federal statutes—was designed toprivilege the status quo.373 By establishing an onerous procedure formaking federal law, the Constitution makes it difficult to enact pre-emptive federal law.374 If so, then the separation-of-powers and feder-alism critiques are subject to a democracy objection. By makingchange so difficult, the Constitution gives a veto power to legislatorsrepresenting a minuscule percentage of the nation’s population, thuspreventing the enactment of preemptive federal law even when a sub-stantial majority of the population favors such legislation. From ademocracy perspective, this would appear to be problematic.

Young would no doubt respond that he is not criticizing the mod-ern position as inconsistent with some ideal of democracy, but ratheras inconsistent with “our” democracy—that is, the version of democ-racy embodied in our Constitution.375 Presumably, then, Young isalso not objecting that the modern position is inconsistent with someidealized version of separation of powers or federalism, but ratherwith “our” separation of powers and federalism. It follows that onecannot assess the modern position except by interpreting our Consti-tution and determining if the modern position is consistent with thatinterpretation. In other words, separation of powers and federalismare not a basis for critiquing the modern position distinct from theclaim that the modern position is inconsistent with the Constitution asproperly construed. And, if that is so, then any separation-of-powersor federalism critique presupposes a theory of constitutionalinterpretation.

Ramsey’s theory of constitutional interpretation is very much atextualist and originalist one. I concluded above that neither the con-

372 See Clark, supra note 24, at 1323–24. Young relies on Clark’s scholarship Rthroughout his piece. See Young, supra note 19, at 371 nn.25–26, 378 n.60, 393 n.142, R402 n.192.373 See Clark, supra note 24, at 1338–46; Young, supra note 19, at 402. R374 For a critique of this view, see Vazquez, supra note 116, at 1602–03. R375 Young criticizes the “internationalist” response to the democracy critique of

the modern position on this ground. See Young, supra note 19, at 399. R



stitutional text nor the Founders’ intent required the revisionist posi-tion favored by Ramsey. Being less of a textualist or originalist, Youngplaces more of an emphasis on constitutional “structure” and “princi-ple,” as well as evolved judicial doctrine. A critique based on structureand principle, however, seems vulnerable to the same objection as onebased on separation of powers and federalism: unless Young means tocriticize the modern position as inconsistent with some idealized ver-sion of constitutional structure or principle, he must link this critiqueto some other accepted indicium of constitutional meaning. If hebases his appeal to structure and principle on the Constitution’s textor the Founders’ intent, his arguments are vulnerable to the sameresponse as Ramsey’s. Young’s critique would appear to be distinctfrom Ramsey’s, then, only insofar as he relies additionally on traditionand evolved judicial doctrine.

I considered the modern position’s conformity with pre-Erie doc-trine above and concluded that there were two contending posi-tions.376 Young stresses that, before Erie, customary international lawwas treated as “general law,” but he recognizes that this regime oper-ated very differently from that of present-day State law.377 As Youngnotes, the “general law” regime produced a remarkable degree of uni-formity because State and federal courts viewed themselves asengaged in a common enterprise and thus tended to rely on eachother’s decisions and arrive at very similar results.378 Young is entirelycorrect in concluding that “[t]he revisionist view that a [S]tate choos-ing to apply customary international norms necessarily incorporatesthose norms as [S]tate law would eliminate that sort of imperative.”379

This was precisely the reason that we concluded in Part I that, asbetween the modern and the revisionist positions, the operative law inthe pre-Erie period resembled the modern position more than therevisionist position.

Of course, Young does not believe that those are our only twochoices. He proposes an intermediate status for customary interna-

376 Post-Erie doctrine would of course be an odd ground on which to defend arevisionist position, since revisionism by definition seeks to overturn entrencheddoctrine.377 See Young, supra note 19, at 466–67. R

378 See id. (“Although each [S]tate had the power to interpret general law by itsown lights . . . both [S]tate and federal participants in that system viewed themselvesas participants in a common enterprise and, accordingly, put a strong emphasis onmutual deference and uniformity.”); see also Fletcher, supra note 124, at 1554, R1562–64, 1572–75 (describing the uniformity experienced by the marine insuranceindustry under the general law regime during the early nineteenth century).379 Young, supra note 19, at 467. R



tional law intended to replicate its pre-Erie status as general law. Asdiscussed in greater detail in the next section, Young proposes thatcustomary international law continue to be treated as general law,applicable in the courts whenever a valid choice-of-law rule calls for itsapplication. Except to the very limited extent that he would recognizea role for federal choice-of-law rules, Young would recognize a Statepower to depart from customary international law,380 and he woulddeny the Supreme Court the power to review State court interpreta-tions of such law, even when a federal choice-of-law rule applies.Although formally this regime may be closer to the pre-Erie generallaw regime, in operation it would suffer from the very problems thathe rightly attributes to the revisionist position. With the loss of a senseof common enterprise, and the explicit recognition of a State powerto depart from customary international law, State courts would be farless likely to follow federal precedents, and State courts and even legis-latures may discourage or even prohibit the application of customaryinternational law, as the citizens of Oklahoma recently voted to do.Thus, we are far more likely to get divergent interpretations of cus-tomary international law and even outright violations of it. Innocencelost cannot be so easily regained.

In the end, Young’s approach is less problematic than Ramsey’sonly insofar as he is willing to accept that federal law sometimesrequires the application of customary international law. Because he iswilling to accept such a federal role in very limited circumstances,however, his proposal is only marginally preferable to Ramsey’s.Moreover, the particular approach that Young proposes for determin-ing the applicability of customary international law in our courts—which relies on choice-of-law rules—would be unhelpful, inapposite,and difficult to apply.

2. Young’s Intermediate Status for Customary International Law

Young proposes that customary international law continue to beregarded as general law.381 He argues at length that this status is com-patible with Erie’s rejection of such status for the common law.382 Thekey to his proposal’s consistency with Erie lies in his recognition thatcustomary international law would be applicable in the courts only

380 See, e.g., id. at 480–81 (“[T]he Charming Betsy principle would bolster a require-ment that [S]tate departures from longstanding customary international norms beindicated by clear legislation or judicial precedents . . . .”); id. at 496–97 (addressingobjection that States may depart from customary international law).381 See id. at 468.382 See id. at 467–74.



insofar as an “otherwise valid” choice-of-law rule calls for its applica-tion to a particular case.383 As Young recognizes, choice-of-law rulesare themselves the creatures of either State or federal law.384 Sincethe applicability of customary international law would thus be deter-mined by State or federal law, Young’s proposal is consistent withErie’s statement that, “[e]xcept in matters governed by the FederalConstitution or by Acts of Congress, the law to be applied in any caseis the law of the State.”385

Young stresses that the applicability of customary internationallaw, in both State and federal courts, will ordinarily be determined byState choice-of-law rules386—presumably the choice-of-law rules of theState in which the court is sitting.387 He acknowledges, however, thatfederal law could play a role in some cases.388 I will first discuss thestatus of customary international law insofar as Young would deter-mine its applicability by reference to State choice-of-law rules. I shallthen discuss its status insofar as Young would permit federal law todetermine its applicability.

a. State Choice-of-Law Rules

Young proposes that customary international law be treated asgeneral law, to be applied by the courts when an otherwise validchoice-of-law rule calls for its application.389 The choice-of-law rulesthat Young has in mind are the rules the courts apply to resolve con-flicts between the laws of different States or nations.390 An initialproblem with this proposal is the uncertainty that it would bring to

383 See id. at 468.384 See id. at 483.385 Erie R.R. Co. v. Tompkins, 304 U.S. 64, 78 (1938); see Young, supra note 19, at R

483. Erie’s list of possible federal sources of law is underinclusive, if only because itexcludes treaties. Cf. U.S. CONST. art. VI, § 2 (stating treaties are the “supreme Law ofthe Land”). The point for now, however, is that Erie appears to hold that all lawapplied in the courts of this country is either State or federal. This requirementwould clearly be satisfied by Young’s proposal if foreign law were regarded as incorpo-rated into the law of the forum State whenever choice-of-law rules call for its applica-tion. See infra note 429 and accompanying text. If one resists that conceptualization, Rthen Erie must be understood to establish instead that all law applied in the courtsmust be applied as a result of authorization by either State or federal law (since theState and federal courts commonly apply foreign law).386 See Young, supra note 19, at 370. R

387 See Klaxon Co. v. Stentor Electric Mfg. Co., 313 U.S. 487, 496 (1941).388 See Young, supra note 19, at 470. R

389 See id. at 370.390 See id. at 470.



the question of the applicability of customary international law.391

The proposal would increase uncertainty in at least two respects: (a)different States approach choice of law differently and (b) many ofthe approaches used are themselves highly indeterminate. A morefundamental problem is that these choice-of-law rules are inapposite,as they address a sort of conflict that is very different from thatbetween customary international law and State law. Since States arefree to adopt special choice-of-law rules for a particular issue, presum-ably most States will end up articulating special rules to determine theapplicability of customary international law. If so, then it is unclearwhat is gained by regarding these rules as choice-of-law rules, asopposed to State rules regarding the incorporation of customaryinternational law as State law.

In any event, to the extent that the matter turns on State rules(whether regarded as choice-of-law rules or rules about incorporationof customary international law as State law), the ultimate determina-tion of whether to apply customary international law will be made bythe States, which would be free to deviate from, or decline altogetherto apply, customary international law. The State would be able to pre-clude such review by making it clear that it is declining to incorporatecustomary international law or (what may amount to the same thing)that its choice-of-law rule precludes its application. Thus, except tothe extent that Young would permit the application of a federalchoice-of-law rule, Young’s proposal suffers from the same basic prob-lem as Ramsey’s.

i. The Diversity and Indeterminacy of Existing Choice-of-Law Approaches

An initial concern with Young’s proposal is the great uncertaintythat would be produced through a resort to State choice-of-law rules.Dean William Prosser once described choice of law as a “dismalswamp,”392 and matters have only gotten worse since he wrote. At onetime, most States applied more or less the same choice-of-law rules,those set forth in the First Restatement.393 The First Restatementapproach often determined the applicable law by reference to thelocation in which a key event occurred. For example, in tort cases, theapplicable law was that of the state in which the injury occurred (the

391 See Daniel J. Meltzer, Customary International Law, Foreign Affairs, and FederalCommon Law, 42 VA. J. INT’L L. 513, 520, 532–36 (2002).392 William L. Prosser, Interstate Publication, 51 MICH. L. REV. 959, 971 (1953).393 RESTATEMENT (FIRST) OF CONFLICT OF LAWS (1934).



lex loci delicti).394 This approach sometimes produced harsh and seem-ingly arbitrary results, but it had the virtue of being relatively determi-nate. Over time, some States came to reject the First Restatementapproach in favor of some version of the interest analysis first pro-posed by Professor Brainerd Currie.395 Later, the Second Restatementarticulated a multifactor balancing approach whose most noteworthyfeature is the high degree of its indeterminacy.396 Today, with theStates employing seven different choice-of-law approaches,397 most ofwhich are themselves highly indeterminate, the choice-of-law situationin the United States is widely regarded as being in a state of crisis.398

Because of the diversity of approaches in use in the United States, theparticular approach that will govern the applicability of customaryinternational law in this country under Young’s proposal will dependon where the suit is brought. Since most of the approaches are them-selves highly indeterminate anyway, the use of such rules will be unap-pealing to anyone who values certainty and predictability in the law.

ii. The Inappositeness of Choice-of-Law Rules

More importantly, the choice-of-law rules to which Young wouldrelegate the question of the applicability of customary internationallaw are inapposite, as they address a completely different type of

394 See id. §§ 378–390.395 See BRAINERD CURRIE, Notes on Methods and Objectives in the Conflict of Laws, in

SELECTED ESSAYS ON CONFLICTS OF LAWS 177, 183–87 (1963).396 See RESTATEMENT (SECOND) OF CONFLICT OF LAWS § 6(2) (1971); Larry Kramer,

Choice of Law in the American Courts in 1990: Trends and Developments, 39 AM. J. COMP. L.465, 466 (1991) (“To be sure, judges make decisions based on unarticulated intuitionin fields other than choice of law, but in no other field do they seem quite so comforta-ble about it. Much of the blame can perhaps be attributed to the dominance of theSecond Restatement, since its undirected, multifactor analysis invites post-hoc ratio-nalizing of intuitions about the applicable law.”); Kermit Roosevelt III, The Myth ofChoice of Law: Rethinking Conflicts, 97 MICH. L. REV. 2448, 2466 (1999) (“[T]he SecondRestatement . . . lists a dizzying number of factors with no hint as to their relativeweight.”).397 See Symeon C. Symeonides, Choice of Law in the American Courts in 2009: Twenty-

Third Annual Survey, 58 AM. J. COMP. L. 227, 231 (2010).398 See, e.g., Michael H. Gottesman, Draining the Dismal Swamp: The Case for Federal

Choice of Law Statutes, 80 GEO. L.J. 1, 1–2 (1991) (advocating congressional action torestore determinacy and uniformity to choice-of-law conflict); Samuel Issacharoff, Set-tled Expectations in a World of Unsettled Law: Choice of Law After the Class Action FairnessAct, 106 COLUM. L. REV. 1839, 1839–40 (2006) (discussing the inadequacy of currentchoice-of-law doctrine, particularly with regard to class action suits); Jed J. Borghei,Note, Class Action Fairness: A Mature Solution to the 23(b)(3) Choice of Law Problem, 95GEO. L.J. 1645, 1645–46 (2007) (arguing for a solution to choice-of-law problems thatrespects federalist and nonfederalist concerns).



choice-of-law problem. These rules were designed for horizontalchoice-of-law questions: a choice between the law of two states, bothtypically addressing the legal relations of private parties. The choice-of-law issue arises because the parties’ transaction or relationshipspanned more than one state or the parties are from different states.The choice-of-law issue is horizontal because the states whose laws arecontending for application are coequal, whether they are States of theUnion or states in the international sense.399 Neither state purportsto impose obligations on the other. If the courts of one state declineto apply the law of the other state to the particular dispute, it is gener-ally on the theory that the law is inapplicable because that state’s con-nection to the dispute is less direct or less relevant than that of theforum state or of a third state. The forum state’s decision not to applyanother state’s law leaves that state’s law untouched in its core applica-tions—disputes in which all of the parties are from the same state andall of the relevant acts took place in that state’s territory. Such a deci-sion merely declines to apply the law to peripheral cases.

A conflict between the law of a State and international law is of acompletely different character. Such a conflict is “vertical” in thesense that one of the two laws purports to impose limits on the other.It is the very point of international law (when it applies) to limit theState’s freedom to, for example, entertain a suit against a diplomat orto execute a minor.400 For a State not to apply that law is to deny theapplicability of the law not at its periphery, but at its core. Unlike aState’s decision not to apply the law of a sister State or of a foreignnation to a particular dispute, a State’s decision not to apply custom-ary international law results in a violation of such law.

Being a vertical conflict, the conflict between State law and inter-national law is analogous to that between State law and federal law.Normally, in the case of a vertical conflict, if the “higher” law is law atall, it will categorically prevail in the event of a conflict. This is theresult for which the Supremacy Clause provides.401 In the event of a

399 A conflict between the law of a State of the Union (say, New York) and a statein the international sense (say, France) might appear not to be a contest between thelaws of coequal sovereigns. But, from the perspective of international law (both pub-lic and private) the law of New York is considered national (or municipal) law, asinternational law does not distinguish between national law and the law of domesticsubunits. Cf. Responsibility of States, supra note 94, at 41. R400 These are two of the examples Young considers. See Young, supra note 12, at

474–81. As to whether there really is a customary international law prohibition of theexecution of minors applicable against the United States, see infra text accompanyingnote 426. R401 If anything, the conflict between State law and international law is more acute

than the sort of conflict usually addressed by the Supremacy Clause. The latter type



conflict between a law that purports to limit what a State may do and aState law that requires or permits what is prohibited, to resolve theconflict by recourse to a choice-of-law rule that permits the applica-tion of the second law is to deny the first law the status of a bindinglaw at all.402

The inappositeness of horizontal choice-of-law rules is well illus-trated by consideration of the rules that were once widely applied toresolve such conflicts and continue to be applied in some States (andin the rest of the world).403 As noted, the choice-of-law rule that onceprevailed for tort cases was the lex loci delicti rule, under which the lawto be applied was the law of the place where the injury occurred.404 InYoung’s example of a tort claim under State law against a foreign dip-lomat entitled to immunity as a matter of international law whereState law does not recognize an immunity,405 the lex loci delictiapproach would require the application of the law of the place wherethe tort occurred. Since international law has no defined territory,perhaps it would never be applicable under a lex loci delicti approach.Alternatively, one could say that the “territory” of international law isthe whole world (or at least those nations of the world bound by therule in question), in which case the lex loci delicti approach would not

of conflict is typically between a federal statute regulating private parties and a Statelaw regulating private parties. The State and federal laws are deemed to be in conflictif they regulate the private parties differently. The Supremacy Clause would be violatedif the State court declined to apply the federal law, but the State would not be violat-ing the federal statute by failing to apply it. In contrast, because international lawapplies to state action and purports to bind the State, the State would be directlyviolating international law if its courts applied a State law that conflicted with it.402 Young recognizes that the conflict between State and federal law is a vertical

one, but he maintains that the conflict between State law and international law ishorizontal because there is no “lexical rule” such as the Supremacy Clause to specify ahierarchy. See Young, supra note 12, at 484. But the Supremacy Clause merely setsforth a rule to resolve the vertical conflict; it is not what makes the conflict a verticalone. The conflict between international law and State law is vertical because interna-tional law purports to limit states, just as federal law purports to limit State law. It isfor the same reason that a conflict between the Constitution and a federal statute is avertical one, even though the Supremacy Clause does not specify a hierarchy asbetween the two. Here, too, it is thought to be self-evident that the law that purportsto restrict the other, if it is law at all, must control. See Marbury v. Madison, 5 U.S. (1Cranch) 137, 180 (1803).403 The following States adhere to the First Restatement approach for torts, con-

tracts, or both: Alabama, Florida (contracts only), Georgia, Kansas, Maryland, NewMexico, North Carolina (torts only), Oklahoma (contracts only), Rhode Island (con-tracts only), South Carolina, Tennessee (contracts only), Virginia, West Virginia (tortsonly), Wyoming. See Symeonides, supra note 397, at 231–32. R404 See supra note 394 and accompanying text. R405 See Young, supra note 19, at 479–81. R



help us decide between the two conflicting laws, since the territory ofthe international law rule (the world) would overlap with that of theState in which the tort occurred. The latter is of course true of con-flicts between State and federal law (that is, the tort occurred in boththe State and the United States), which is why territorial choice-of-lawrules are unhelpful in resolving such conflicts.

Young’s proposal is less implausible if one uses one of the mod-ern approaches to choice of law that some States have adopted, butthese newer approaches too are addressed to horizontal conflicts andare inapposite to vertical conflicts such as the one under discussion.The choice-of-law revolution’s first significant incursion came in theform of interest analysis. The one feature of Currie’s interest analysisthat is widely regarded as an advance is its identification of “false con-flicts,” that is, situations in which only one of the States whose conflict-ing laws are contending for application has an interest in applying itslaw to the particular case.406 For example, if two citizens of New Yorkare riding in an automobile and have an accident in Ontario, andOntario law denies the passenger a right to recover against the driverfor negligence while New York law permits recovery, only New York(the State of the common domicile) has an interest in applying its lawto govern the liability of the driver to his guest.407 The purpose ofOntario’s law is to protect hosts from the ungratefulness of theirguests, and Ontario has no interest in applying its law because thehost in this case was not from Ontario. Despite the complexities thatmay arise in determining the interests underlying the contendinglaws, interest analysis has been widely regarded as a success insofar asit has facilitated the resolution of such false conflicts.408 However,conflicts between State law and international law will never be falseconflicts. As noted above, the whole point of international law is toconstrain states. Thus, if a State law conflicts with international law,the conflict will always be a true one.409

406 See LEA BRILMAYER & JACK GOLDSMITH, CONFLICTS OF LAWS 218–19 (5th ed.2002); William F. Baxter, Choice of Law and the Federal System, 16 STAN. L. REV. 1, 8(1963).407 These were the facts of Babcock v. Jackson, 191 N.E.2d 279 (N.Y. 1963), in which

the New York courts first adopted interest analysis. See id. at 280.408 But see Lea Brilmayer, Rights, Fairness, and Choice of Law, 98 YALE L.J. 1277,

1300–01 (1989).409 This is true even if the suit involves a dispute between foreign parties in which

all the events took place in a foreign country. In such a case, the relevant conflictwould be between the law of the country where the conduct took place and interna-tional law. If there is a conflict at all, it is because the law of that foreign country doesnot accord with international law. Since it is the point of international law to con-strain that country, the conflict will always be a true one.



Young argues that the conflict between a State law that recognizesa tort cause of action but is silent regarding immunity and a rule ofinternational law that entitles the defendant to immunity would be afalse conflict.410 But this misconceives the concept of a false conflict.Before asking whether a conflict is true or false, one must determinewhether the laws contending for application conflict. To determinewhether they conflict, one must determine the content of the laws inquestion. In the case of a State whose law recognizes a generally appli-cable tort cause of action but is silent as to immunity, the court firsthas to decide whether the State’s law implicitly recognizes an immu-nity. This is a matter of interpreting the State’s domestic law. If theState recognizes an immunity, then there is no conflict with interna-tional law and thus no choice-of-law issue to resolve. If the Statedenies an immunity, then there is a conflict. Only then does oneengage in interest analysis to determine if the conflict is true or false.Whether the conflict is true depends on whether the State’s interestswould be advanced by having its law applied to the case, which in turndepends on such factors as whether the plaintiff is a citizen of thatState. If the plaintiff is a citizen, and if we assume that the State’scause of action was designed to benefit citizens,411 then the conflictwould be a true one.412

Young posits a tort committed by a diplomat in Massachusettsinjuring a Massachusetts resident.413 If Massachusetts law deniedimmunity and international law conferred one, there would be a con-flict. If Massachusetts denied immunity in order to protect its citizenswho are injured by diplomats, the conflict would be a true one(because the injured party was from Massachusetts). On the otherhand, if Massachusetts law were construed to grant immunity, therewould be no conflict at all because State law and international law arethe same. What Young considers a false conflict is in fact a nonexis-tent conflict. The State law’s silence on the question of immunityraises an issue of interpretation, not a choice-of-law issue. A State

410 Young says that such a conflict would not be an “actual” conflict, but presuma-bly he is invoking Currie’s concept of a true conflict (the opposite of a false conflict),as he cites Larry Kramer’s discussion of Currie’s false conflict analysis. See Young,supra note 19, at 469 (citing Larry Kramer, Rethinking Choice of Law, 90 COLUM. L. REV. R277, 311 (1990)).411 This is an assumption often made in interest analysis, although scholars have

criticized it. See Brilmayer, supra note 408, at 1300. R412 If the plaintiff is not from the forum State, then the relevant conflict will be

between the law of the State of the plaintiff’s domicile and international law. See supranote 409 and accompanying text. If the law of that State grants immunity, there is no Rconflict. If the law of that State denies immunity, the conflict will be a true one.413 See Young, supra note 19, at 479. R



might resolve that question through a background rule that all that isnot explicitly prohibited is permitted. Young himself proposes thatState and federal courts resolve the issue by assuming that the Stateshad retained the background understanding, expressed by Blackstone(among others), that the law of nations was part of English commonlaw.414 Under this approach, State law would be understood to incor-porate customary international law unless the legislature has expresslyrejected it. That is certainly a reasonable solution, but it is not achoice-of-law solution.415

Resolving true conflicts has been the bete noire of interest analysis.Currie originally proposed always applying forum law.416 Thisapproach would, of course, always reject the application of interna-tional law. Most States have rejected the forum law solution.417 Cali-fornia has adopted Professor William Baxter’s comparativeimpairment approach, under which true conflicts would be resolvedby applying the law that would be impaired most if not applied,418 anotoriously difficult test to apply. In the case of a vertical true con-flict, it would seem that the higher law would always be completelyimpaired if not applied, whereas, depending on how one character-ized its purpose, State law may or may not be fully impaired if notapplied insofar as it violates international law. This test would thusseem to favor the application of international law.

Other States have adopted Dean Robert Leflar’s “better law”approach.419 This test differs from the others in that it openly

414 See id. at 480.415 At least, it is not a horizontal choice of law solution. It might be understood as

a vertical choice-of-law solution, but, as such, it would not depend on horizontalchoice-of-law concepts such as that of a false conflict.416 See CURRIE, supra note 395, at 184; Brainerd Currie, Comments on Babcock v. R

Jackson, a Recent Development in Conflict of Laws, 63 COLUM. L. REV. 1212, 1237–38,1242 (1963). The discussion in the remainder of this subsection assumes that the lawof the State conflicts with customary international law – in other words, in Young’shypothetical diplomatic immunity case, that customary international law entitles thediplomat to immunity and that Massachusetts law does not recognize an immunity fordiplomats, either by incorporating customary international law or independently. Asnoted, if the plaintiff is a citizen of Massachusetts, the conflict will be a true one.417 Two States—Kentucky and Michigan—follow the forum law approach, but for

torts only. See Symeonides, supra note 397, at 231, 248 n.80. R418 See Bernhard v. Harrah’s Club, 546 P.2d 719, 723–26 (Cal. 1976).419 See Robert A. Leflar, Choice-Influencing Considerations in Conflicts Law, 41 N.Y.U.

L. REV. 267, 295–304 (1966); Robert A. Leflar, Conflicts Law: More on Choice-InfluencingConsiderations, 54 CALIF. L. REV. 1584, 1588, 1590–93 (1966). The following Statesapply the “better law” approach: Arkansas (torts only), Minnesota, New Hampshire(torts only), Rhode Island (torts only), and Wisconsin. See Symeonides, supra note397, at 231–32. R



requires an evaluation of the substance of the contending laws. Anapproach that leaves it to State courts to evaluate the merits of a ruleof international law that binds the United States (and which the fed-eral government thus very likely agrees with, at least tacitly) is highlyproblematic. The structural problem with leaving compliance withinternational law to the States has already been noted. Leaving it toState judges to assess the desirability of a given norm of customaryinternational law seems particularly inadvisable, as State judges wouldappear to be even less qualified than State executives or legislatures toevaluate norms of international law.

The largest number of States (though barely a majority) followsthe approach of the Restatement (Second) of the Conflict of Laws.420 TheSecond Restatement takes a hybrid approach, designating the Statewhere a particular act took place as the State whose law presumptivelyapplies, but permitting that law to be displaced if another State’sinterest is greater.421 To guide the determination of the latter ques-tion, the Second Restatement sets forth multiple factors to be consid-ered.422 As applied to a vertical conflict of the sort we areconsidering, the initial presumption will be useless if, as discussedabove, the “territory” of international law is the world (or all statesbound by the relevant international law norm).

The analysis provided in the Second Restatement to displace thepresumptively applicable law is notoriously indeterminate. TheRestatement sets forth an unexhaustive list of seven factors to be consid-ered.423 The upshot is that the State courts will be deciding whetherinternational law is to be applied based on their case-specific conclu-sions about whether applying such law would be reasonable. It is thusnot surprising that Young believes that “[his] approach’s heavy reli-

420 The following States follow the Second Restatement: Alaska, Arizona, Colorado,Connecticut, Delaware, Florida (torts only), Idaho, Illinois, Iowa, Kentucky (contractsonly), Maine, Michigan (contracts only), Mississippi, Missouri, Montana, Nebraska,Nevada (torts only), New Hampshire (contracts only), New Jersey (torts only), Ohio,Oklahoma (torts only), South Dakota, Tennessee (torts only), Texas, Utah, Vermont,Washington, and West Virginia (contracts only). See Symeonides, supra note 397, at R231–32.421 See RESTATEMENT (SECOND) OF CONFLICT OF LAWS § 145 cmt. d (1971).422 See id. § 145(2).423 See id. § 6(2) (stating that the factors to be considered “include (a) the needs

of the interstate and international systems, (b) the relevant policies of the forum, (c)the relevant policies of other interested states and the relative interests of those statesin the determination of the particular issue, (d) the protection of justified expecta-tions, (e) the basic policies underlying the particular field of law, (f) certainty, pre-dictability and uniformity of result, and (g) ease in the determination and applicationof the law to be applied”).



ance on choice of law rules [means] that those rules should generallyshake out in a way that conforms to our intuitive notions of wheninternational law ought to govern and when it should not.”424 Thiswould certainly be true if the court applied the Second Restatementapproach, or most versions of interest analysis, as those approacheswould essentially license courts to decide the applicable law in accor-dance with their intuitions about the desirability of the particularrules of international law involved, or their more general views aboutwhether the matter is one that international law “ought to govern.”For example, Young’s application of the Second Restatement test to aconflict between a State’s law that would subject a minor to the deathpenalty and a rule of customary international law that prohibits theapplication of the death penalty to minors seems to reflect Young’sown intuition that this is a matter best left to local decisionmaking.“If, after years of moral and constitutional debate, we cannot say thatcapital punishment is inconsistent with ‘evolving standards ofdecency’ under the Eighth Amendment, then it would be shocking tosee the controversy resolved through the deus ex machina of interna-tional law.”425 In this particular case, Young’s intuition appears toaccord with that of the rest of the world, which recognizes that statesthat have persistently objected to an evolving rule of customary inter-national law are exempt from the rule if it does eventually developinto customary law.426 But, if applied to a rule of customary law thatactually binds the United States, and with which the federal govern-

424 Young, supra note 19, at 474. R425 Id. at 477.426 See RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW § 102 cmt. d (1987)

(“[I]n principle a state that indicates its dissent from a practice while the law is still inthe process of development is not bound by that rule even after it matures.”); CurtisA. Bradley, The Juvenile Death Penalty and International Law, 52 DUKE L.J. 485, 492, 539(2002) (concluding that international law arguments made against the juvenile deathpenalty have “significant weaknesses”, that “the United States has persistently objectedto—and thereby legally opted out of—any customary international law restriction onthe juvenile death penalty,” and that “the jus cogens challenge to the juvenile deathpenalty seems unpersuasive”). But cf. Connie de la Vega & Jennifer Brown, Can aUnited States Treaty Reservation Provide a Sanctuary for the Juvenile Death Penalty?, 32 U.S.F.L. REV. 735, 770 (1998) (“[T]he prohibition of the juvenile death penalty is not onlycustomary international law, but also jus cogens, and a norm to which the renegadeUnited States should be bound.”); Joan F. Hartman, “Unusual” Punishment: The Domes-tic Effects of International Norms Restricting the Application of the Death Penalty, 52 U. CIN.L. REV. 655, 686 (1983) (“Mere failure to conform to the norm is not tantamount to asubstantive protest.”); Lynn Loschin, The Persistent Objector and Customary Human RightsLaw: A Proposed Analytical Framework, 2 U.C. DAVIS J. INT’L L. & POL’Y 147, 171 (1996)(“Passive failure to bring domestic law into conformity is not sufficient, particularlywhen the objecting state has had multiple opportunities to discuss the issue and make



ment thus at least tacitly agrees, an approach that calls on State judgesto decide whether to apply the law on the basis of their judgmentabout whether the rule is a good one or whether international law“ought to govern” is highly problematic. If the applicability of inter-national law in our courts is to be decided on this basis, the States—and in particular State judges—would seem to be the wrong organs tomake that determination.

iii. The Likelihood of Special Choice-of-Law Rules

To the extent that Young would determine the applicability ofcustomary international law by reference to State choice-of-law rules,he would leave it to rules that States are free to alter as they wish.There is of course nothing to prevent States from changing their gen-eral approach to choice of law or from adopting a different approachfor conflicts involving international law than for the sorts of horizon-tal conflicts that they are most accustomed to confronting. (Depecageis a familiar technique in choice of law.427) The most likely to adopt aspecial choice-of-law rule for conflicts involving international law arethe States that adhere to the First Restatement’s territorial approach,since, as we have seen, those approaches are very likely to be useless inthe case of vertical conflicts. The more modern approaches to choiceof law will be of any use only to the extent that they are indeterminate.But these approaches too are fraught with problems and thus likely tobe displaced with special choice of law rules.

If a special choice-of-law rule were adopted, it would likely be aflat rule declaring one or the other to be supreme unless the legisla-ture specifically provided otherwise. That is of course the sort of rulethat the Constitution adopts to resolve the analogous conflict betweenState and federal law. It is also the sort of rule that Oklahoma chosein its recent constitutional amendment.428 In the case of vertical con-flicts, this sort of rule would appear to be the only one available. A

principled objections. Therefore, . . . the analysis would . . . deny the United Statespersistent objector status.” (footnote omitted)).427 Depecage is defined as “[a] court’s application of different state laws to differ-

ent issues in a legal dispute; choice of law on an issue-by-issue basis.” BLACK’S LAW

DICTIONARY 502 (9th ed. 2009). The Restatement embraces depecage. See, e.g.,RESTATEMENT (SECOND) OF CONFLICT OF LAWS § 7(3) (“The classification and inter-pretation of local law concepts and terms are determined in accordance with the lawthat governs the issue involved.”).428 The Oklahoma constitutional amendment goes further than a rule of State law

superiority, as it would bar application of international law even if it does not conflictwith State law. But this rule could be conceived as a rule of State law superiority,coupled with a stipulation that State law is comprehensive.



rule of State law superiority would amount to a rule denying custom-ary international law the status of law, although such a rule could besupplemented (as a matter of State law) by a Charming Betsy presump-tion, under which State law will be interpreted to be consistent withinternational law if ambiguous. Alternatively, the State could deter-mine that particular rules of customary international law but notothers shall be applicable in the State, depending on the State’s viewsabout their desirability. The “choice of law” label seems entirely inaptto describe an approach of this sort. A State that picks and choosesnorms of international law in this way would appear to be engaged ina process of selective incorporation of international law.429

iv. The Role of the Federal Courts

Since customary international law would have the status of “gen-eral law” under Young’s proposal, Young concludes that cases turningon such law would not “arise under” federal law for purposes of Arti-cle III.430 The federal courts would nevertheless have jurisdiction overcases involving customary international law if the parties werediverse.431 In such cases, Young says that he would permit the federalcourts to use their own judgment as to the content of customary inter-national law, as they did before Erie.432 It would seem to follow from

429 Even in the horizontal choice-of-law context, a State’s decision that anotherState’s law is applicable is often conceptualized as the first State’s incorporation of theother State’s law as its own. See Guinness v. Miller, 291 F. 769, 770 (S.D.N.Y. 1923)(Hand, J.), aff’d sub nom. Guiness v. Miller, 299 F. 538 (2d Cir. 1924), aff’d in part, rev’din part sub nom. Hicks v. Guinness, 269 U.S. 71 (1925); WALTER WHEELER COOK, THE

LOGICAL AND LEGAL BASES OF THE CONFLICT OF LAWS 20–21 (1942). This conceptual-ization seems even more apt in the vertical choice-of-law context, at least where, asYoung suggests, the “choice” to apply international law is left to the States.430 See Young, supra note 19, at 468. This assumes rejection of theories of protec- R

tive jurisdiction such as Professor Wechsler’s, under which Congress would be empow-ered to grant the federal courts “arising under” jurisdiction if it would have had thepower to displace State law with a federal rule, but without actually displacing Statelaw. See Wechsler, supra note 50, at 224–25. See generally Vazquez, supra note 50 (dis- Rcussing various theories of protective jurisdiction). Under this theory, Congress couldconfer jurisdiction over cases arising under customary international law because Con-gress has the power to displace State law with customary international law under theOffenses Clause. Young rejects protective jurisdiction. See Young, supra note 49. In Rany event, as discussed above, the Supreme Court’s jurisdiction over a case and itspower to displace a State court’s interpretation of customary international law areseparate questions. See supra text accompanying note 66. R431 Presumably such jurisdiction would also exist if a customary international law

issue arose in a claim that is ancillary to a claim that arises under a federal statute ortreaty or the Constitution.432 See Young, supra note 19, at 471–72. R



Young’s theory, however, that the federal courts would be free to usetheir own judgment about the content of customary international lawonly in limited circumstances. In the absence of a federal choice-of-law rule, the federal courts would have to apply the choice-of-law rulesof the State in which it sits. If the applicability of customary interna-tional law were a matter of State choice of law, then a federal courtwould not be free to apply customary international law if the State inwhich it sits has rejected the application of such law, as the citizens ofOklahoma have done. Additionally, if the State’s “choice” to applycustomary international law were expressly made contingent on a par-ticular interpretation of such law, the federal court would appear tobe bound by that interpretation. At the other end of the spectrum,the federal courts would be able to apply their own interpretation ofcustomary international law if the State has made clear that it is incor-porating or has chosen to apply the correct interpretation of such law(and if the State courts have not yet interpreted the particular norminvolved). In my view, the federal courts could also defensibly applytheir independent judgment regarding the content of the interna-tional law norm if the State has not specified a special choice-of-lawrule for conflicts between State law and customary international law.The rationale would be that (a) horizontal choice-of-law rules areinapposite and (b) presumptive incorporation of customary interna-tional law as State law can be assumed, in the absence of explicit Staterejection of such law, on the theory that such law was incorporated asEnglish common law at the time of the Founding. Even in such cir-cumstances, however, Young’s theory appears to require the conclu-sion that the States would be free to reject an interpretation ofcustomary international law by the federal courts, even the SupremeCourt, either by legislation or judicial decision, and once the State hasdone so, the federal courts would thereafter be bound to follow thestate courts’ interpretation.433

433 Young suggests that, if a basis of jurisdiction other than “arising under” juris-diction existed (for example, diversity), the Supreme Court could take cases to reviewissues of customary international law. See id. at 500. Presumably, Young means thatthe Supreme Court can review such cases when they arise in the lower federal courts,as the Supreme Court lacks jurisdiction to review cases from the State courts on thebasis of diversity. See 28 U.S.C. § 1257 (2006). If Young does have Supreme Courtreview of State court decisions in mind, Congress would have to pass a statute author-izing such review on the basis of diversity. But if Congress were concerned enoughabout customary international law to warrant such a statute, it would presumably beconcerned enough about State interpretations of customary international law todeclare such law (or perhaps certain norms of customary international law, or normsmeeting certain requirements of clarity or acceptance) to be preemptive federal law.Congress would clearly have the power to do so under the Offenses Clause, and this



v. Summary

Young’s proposal to employ State choice-of-law rules to deter-mine the applicability of customary international law is problematicbecause States employ a variety of different approaches to choice oflaw and most of the States’ approaches are highly indeterminate.More fundamentally, such rules are inapposite, as they address hori-zontal choice-of-law problems. The determinate approaches are thuslikely to be useless; the indeterminate ones less so, but only becausethey essentially leave the question to the unguided discretion of thelower courts. In any event, States would be free to adopt specialchoice-of-law rules for conflicts involving international law and wouldbe well advised to do so, given the problems with applying horizontalchoice-of-law rules. If a State does adopt a special rule to address theapplicability of customary international law, it may as well frame therule as one about the incorporation of customary international law asState law. There would appear to be nothing to gain from employingthe choice-of-law terminology, and much confusion to be avoided byabandoning it.

The federal courts could defensibly presume that a State’s ruleregarding the applicability of customary international law is that ofthe Charming Betsy: such law applies unless the legislature has specifi-cally rejected it. But, if customary international law truly were applica-ble only through otherwise valid choice-of-law rules, then, to theextent the choice-of-law question were governed by State law, theStates would retain the power to reject its application. States would befree to adopt a rule like that approved by the citizens of Oklahomaunder which customary international law never applies, and theycould even adopt a rule under which ambiguous State laws will beconstrued whenever possible so as to violate international law. Thus,except to the limited extent that he would recognize federal choice-of-law rules, Young’s approach suffers from the same problems as Ram-sey’s. It merely uses a different vocabulary to avert violations of inter-national law in the absence of clear State decisions to depart fromsuch law.

approach would give the Court power to review cases without regard to diversity whileobviating “protective jurisdiction” issues under Article III. That Young is not maintain-ing that the Supreme Court would be able to review a State court’s interpretation ofcustomary international law even in cases falling within Article III’s diversity clauses isshown by his statement, discussed in the next subsection, that State courts wouldretain ultimate authority over the interpretation of customary international law evenwhen federal law requires its application. See Young, supra note 19, at 509. R



b. Federal Choice-of-Law Rules

Young’s approach differs from Ramsey’s in that he would recog-nize certain circumstances in which the applicability of customaryinternational law would be determined by federal law. Young prefersto characterize the role of federal law here as a “federal common lawconstraint on State choice of law.”434 But it would appear to make nodifference whether it is characterized that way or instead as a federalchoice-of-law rule that displaces the State choice-of-law rule or,indeed, as the federal incorporation of a particular rule of customaryinternational law displacing a contrary State law. However the federalrole is characterized, the result is the same: federal law requires theapplication of customary international law to the dispute.

The irrelevance of the characterization is shown by Young’sexample of a case in which a court viewed federal law as (possibly)imposing a limit on State choice of law: Chief Judge Hand’s opinionin Bergman v. De Sieyes.435 As discussed above, Hand in Bergman statedthat New York law was applicable to the question of a foreign diplo-mat’s immunity, but acknowledged that “an avowed refusal to accept awell-established doctrine of international law, or a plain misapprehen-sion of it, [might] present a federal question.”436 While Youngregards the role of federal law in the situation described (and leftopen) by Hand as imposing a limit on State choice of law, it couldequally (or more accurately) be characterized as “choosing” interna-tional law as the applicable law, or even as incorporating a subset ofcustomary international law as preemptive federal law.437 Indeed,Hand seems to have had the last of these conceptualizations in mind,as he wrote that a “federal question” might be presented in the cir-cumstances he was discussing.

Young does not tell us much about the circumstances in which hewould regard it as appropriate for federal law to require application ofcustomary international law, but he does tell us that he regards thisoutcome as appropriate only in rare circumstances.438 His concessionthat federal law might sometimes limit the States’ freedom not tochoose international law comes in his discussion of a diplomat’s claim

434 Young, supra note 19, at 480. R435 See id. at 458, n.479.436 See Bergman v. De Sieyes, 170 F.2d 360, 361 (2d Cir. 1948)437 See Meltzer, supra note 391, at 531 (“[I]nstead of arguing as [Young] does that R

federal interests shape choice of law rules that sometimes call for application of [cus-tomary international law] (for example, of head of state immunity), why not cut outthe middleman, and say that federal interests call for application of a substantive ruleof immunity as a matter of federal common law?”).438 See Young, supra note 19, at 508–09. R



of immunity from a State law claim, and his analysis suggests that thepossible federal displacement of State choice-of-law rules would haveto occur on a retail basis—that is, with respect to particular rules ofinternational law.439 In this respect, his approach differs from the onesuggested (but not ruled upon) by Hand, which would apparentlyhave required compliance, as a matter of federal law, with (at least) all“well-established” rules of international law. Young says that any judi-cially recognized federal rule requiring compliance with customaryinternational law must be “tied closely . . . to specific concerns aboutpreserving political branch control over foreign policy,”440 as the Sab-batino441 rule was, and he insists that State rules must apply “unlessthose rules pose a significant conflict with specific federal interests,”442

which is the approach to federal common law contemplated by UnitedStates v. Kimbell Foods, Inc.,443 a case not involving customary interna-tional law.444 Young’s treatment of Sabbatino as an outer limit, ratherthan as an example of a more general principle, is questionable giventhe Court’s indication that the federal common law rule that it articu-lated and Jessup’s view concerning the status of customary interna-tional law were both instantiations of the more general propositionthat “legal problems affecting international relations” were a federalmatter.445 Young’s insistence on application of the Kimbell Foodsframework for articulating federal common law overlooks his owninsight that the modern position, or some variant of it, might be justi-fiable outside that framework.446 As discussed in Part I and by Belliaand Clark, the view that customary international law, or some subsetthereof, generally has the force of preemptive federal law can be justi-fied as necessary to preserve federal political branch control over for-

439 See Young, supra note 34, at 29 (outlining this point more clearly and arguing Rthat the Supreme Court embraced this view in its decision in Sosa).440 Young, supra note 19, at 509. R

441 Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398 (1964).442 Young, supra note 19, at 509. R

443 440 U.S. 715 (1979). Bradley, Goldsmith, and Moore take a similar position.See Bradley, Goldsmith & Moore, supra note 33, at 935–36. R

444 See Young, supra note 19, at 509; see also Ernest A. Young, Historical Practice and Rthe Contemporary Debate over Customary International Law, 109 COLUM. L. REV. SIDEBAR

31, 39–40 (2009), www.columbialawreview.org/Sidebar/volume/109/31_Young.pdf.445 See Sabbatino, 376 U.S. at 425.446 Cf. Young, supra note 19, at 447 (“The modern position is not simply a particu- R

lar application of established principles governing federal common law generally, butrather a sui generis regime that must be justified—if it can be justified—outside thatframework.”).



eign policy. The constitutional structure does not require the retailincorporation of customary international law.447

Because Young would regard customary international law as “gen-eral law” even when federal law requires its application, he maintainsthat, even then, the Supreme Court would lack jurisdiction to reviewState court applications of such law, and hence “a [S]tate court wouldretain independent interpretive authority over that law.”448 If so, thenthere would be a significant difference between a federal choice-of-lawrule requiring application of customary international law and a fed-eral incorporation of customary international law as federal law, asclearly the Supreme Court would have jurisdiction to review a Statecourt’s interpretation of customary international law in the latter situ-ation. But it seems to me that, precisely because there is no real dif-ference between the two, the Supreme Court would in fact havejurisdiction to review a State court’s interpretation of customary inter-national law in the former situation. Perhaps the Court would lackthe power to review a State court’s interpretation of such law if theonly point of the federal choice-of-law rule was to require the States togo through the motions of applying customary international law. But,if the federal government had an interest in having international lawapplied, its interest would presumably not be satisfied by a Statecourt’s application of some zany interpretation of it. The reasons thatYoung cites as justifying the adoption of a federal common law choice-of-law rule—such as the “threat to the ability of the federal politicalbranches to conduct foreign affairs”449 that would be posed by thecourt’s failure to apply international law in a given case—would beequally impaired if the State court misapplied such law. Judge Handappeared to believe that federal jurisdiction would exist if federal lawrequired the States to give effect to “well-established” principles ofinternational law. I agree that the Supreme Court would possess juris-

447 Nor does Sosa embrace a “retail incorporation” approach. As discussed above,the issue in Sosa was whether to recognize a private right of action, not whether cus-tomary international law had the status of federal law. (Indeed, Young himself arguesthat Sosa may well have created a federal right of action for damages for violations of ageneral law rule. See Young, supra note 34, at 31.) Moreover, the Court’s limitation of Rthe right of action to certain rules of customary international law does not reflect anadoption of a retail approach (that is, an approach that applies to individual rules ofcustomary international law). The Court, rather, created a private right of action forall rules meeting certain generally applicable requirements. See supra text accompa-nying note 241. The Court did suggest that it might be appropriate to consider cer- Rtain case-specific factors, but only to determine whether the action should be allowedto proceed. See Sosa v. Alvarez-Machain, 542 U.S. 692, 733 n.21 (2004).448 Young, supra note 19, at 509. R449 Id. at 480.



diction to review a State court’s application of customary internationallaw in such circumstances. In other words, if federal law (howeverdenominated) requires application of a norm of customary interna-tional law, the norm is no longer merely “general law.”

If Young were correct about the lack of Supreme Court power toreview State court interpretations of customary international law, thiswould count as a substantial point against his approach, both becauseof the resulting lack of uniformity in the interpretations of interna-tional law within the United States and because of the actual violationsof international law that would occur. Because the Supreme Courtwould in fact have the power to review State court applications of cus-tomary international law when federal law required its application,Young’s approach would actually be less problematic than Ramsey’s.This is so because, despite his denial, Young would in fact recognizesome preemptive force to customary international law. In my view, heis right to do so, but he does not go far enough. His own structuraland doctrinal arguments would support a broader preemptive forcefor such law than he would recognize.

C. Aleinikoff’s Position

Dean Aleinikoff is drawn to an intermediate status for customaryinternational law, but he concludes that the particular approacheschampioned by Young and Ramsey fail on their own terms. Hebelieves that Ramsey’s thesis is internally inconsistent and that Young“uses the apparatus of choice of law to rename well-established normsof preemption and supremacy.”450 Aleinikoff proposes instead thatwe treat customary international law as “nonpreemptive, nonfederallaw.”451 Since this characterization would seemingly regard customaryinternational law as even less federal than Ramsey’s approach (whichpurports to treat such law as nonpreemptive federal law), one mightthink that Aleinikoff’s proposal poses the same problems as Ramsey’sand then some. But, on closer inspection, Aleinikoff’s proposal posesproblems of an entirely different sort.

For Ramsey, the “nonpreemptive” nature of customary interna-tional law meant that it could not be applied if it conflicted with Statelaw. In the face of such a conflict, neither State nor federal courtswould be permitted to apply it. Aleinikoff, on the other hand, viewshis concept of nonpreemptive nonfederal law as being applicable inthe federal courts regardless of what State law has to say. ForAleinikoff, customary international law is nonpreemptive only in the

450 Aleinikoff, supra note 20, at 96. R451 Id. at 97.



sense that it is not applicable in the State courts in the event of aconflict with State law. In other words, customary international law,according to Aleinikoff, “binds federal courts but not state courts.”452

An initial problem with this proposal is that it would reproducethe one aspect of the pre-Erie regime that everyone had thought Eriehad completely ruled out. Under Aleinikoff’s proposal, the lawapplied in the State courts would differ from that applied in the fed-eral courts. The applicable law would thus turn on “the accident ofdiversity.”453 (Aleinikoff makes clear that, under his proposal, custom-ary international law would not be federal law for purposes of arisingunder jurisdiction.454)

Aleinikoff appears to regard his regime as being one in whichcustomary international law, as interpreted by federal courts, wouldeffectively “announce the rule for the federal branches,”455 whereasthe applicability of customary international law to State actors “wouldbe a matter of [S]tate law as to whether and how international law . . .ought to be incorporated.”456 But if his proposal is that customaryinternational law would be applicable in federal court but not in Statecourt (unless incorporated by State law), then it would not alwaysapply to federal officials, and it might sometimes apply to the actionsof State officials even when State law has not incorporated it. Imaginea suit against a federal official seeking to enjoin conduct that allegedlycontravenes customary international law. If customary internationallaw were nonfederal law, such a suit would have to be brought in Statecourt in the absence of diversity.457 The federal official would have aright to remove the case to federal court if he raised a federaldefense.458 Presumably his defense in the typical case would be that afederal statute authorized his conduct. Customary international lawwould then be invoked by the plaintiff to defeat the defense of justifi-cation.459 But the official would not be required to remove the case. If

452 Id. at 99.453 See Klaxon Co. v. Stentor Electric Mfg. Co., 313 U.S. 487, 496 (1941).454 See Aleinikoff, supra note 20, at 98. R455 See id. at 97.456 See id. at 98.457 Cf. Westfall Act, Pub. L. No. 100-694, § 5, 102 Stat. 4563, 4564 (1988) (codified

at 28 U.S.C. § 2679(b)(1) (2006)) (“Any other civil action or proceeding for moneydamages arising out of or relating to the same subject matter against the employee orthe employee’s estate is precluded without regard to when the act or omissionoccurred.” (emphasis added)).458 See Mesa v. California, 489 U.S. 121, 129 (1989) (citing 28 U.S.C. § 1442(a)

(1988)).459 One possible problem with this argument is that customary international law,

even under the modern position, does not necessarily trump federal statutes. See



a federal court but not a State court would apply customary interna-tional law, the officer might well prefer to stay in State court. If theState court declined to apply customary international law and ruled infavor of the officer, Aleinikoff would apparently not permit SupremeCourt review, as the petitioner’s claim would be that the officer vio-lated customary international law, and such law (in Aleinikoff’s view)is nonfederal.460

Conversely, consider an action against a State official allegingthat State conduct violates customary international law. Such anaction could be commenced in federal court if the plaintiff were froma different State, or if the claim under customary international lawwere pendent to a federal claim. Under Pennhurst State School & Hos-pital v. Hoffman (Pennhurst II),461 an action seeking to enjoin such con-duct would have to be dismissed on Eleventh Amendment grounds,462

but courts have held that the Eleventh Amendment does not bar dam-age suits in federal court against State officials for violations of Statelaw.463 Moreover, the compatibility of State law with customary inter-national law could arise in an action between private parties, and ifthe parties were diverse, the issue could be adjudicated in federal

RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW § 115 reporters’ note 4 (“Sinceinternational customary law and an international agreement have equal authority ininternational law, and both are law of the United States, arguably later customary lawshould be given effect as law of the United States, even in the face of an earlier law oragreement, just as a later international agreement of the United States is given effectin the face of an earlier law or agreement. But customary law is made by practice,consent, or acquiescence of the United States, often acting through the President,and it has been argued that the sole act of the President ought not to prevail over alaw of the United States.” (citations omitted)). But Aleinikoff would permit later-in-time custom to prevail over federal statues. See Aleinikoff, supra note 20, at 100. In Rthe more common case of earlier-in-time custom, Aleinikoff is presumably contem-plating its application via the Charming Betsy presumption. Cf. id. at 99.460 If the State court were likely to apply customary international law to the case,

then the officer would presumably remove the case to federal court. If the likelihoodthat the State court would apply international law were uncertain, the officer mightdecide to take his chances in State court. If the State court did ultimately apply cus-tomary international law but ruled against the officer, Aleinikoff would apparently notpermit an appeal to the Supreme Court.461 465 U.S. 89 (1984).462 See id. at 101–02.463 See Pena v. Gardner, 976 F.2d 469, 473–74 (9th Cir. 1992); see also Pennhurst II,

465 U.S. at 111 n.21 (clarifying that Larson v. Domestic & Foreign Commerce Corp.,337 U.S. 682 (1949), held that cases seeking damages against individual officers donot necessarily implicate relief against the sovereign); PAUL M. BATOR ET AL., HART

AND WECHSLER’S THE FEDERAL COURTS AND THE FEDERAL SYSTEM 1203 (3d ed. 1988)(noting that Pennhurst II left open the issue of whether State officials could be sued infederal court for damages resulting from State law violations).



court. In addition, in a case challenging State law as a violation of afederal statute or constitutional provision, a claim under customaryinternational law could be added as a pendent claim, or customaryinternational law could be brought in by way of the Charming Betsypresumption to aid in the application of the relevant federal law, asAleinikoff proposes.464 Thus, if customary international law wereapplicable in federal court even when it conflicts with State law, itwould control the validity of State action in some circumstances.

For this reason, Aleinikoff’s statement that customary interna-tional law would be “nonpreemptive” is inaccurate. Customary inter-national law would serve to preempt State law in cases in which federaljurisdiction could be established on the basis of diversity of citizenshipor when the customary law claim is pendent to a federal claim. Thus,he is wrong to claim that his proposal does not “rais[e] serious feder-alism issues.”465 Admittedly, such issues would arise only in somecases, probably few—but that itself is a significant problem with theproposal, as the applicability of customary international law woulddepend on the fortuity of federal jurisdiction.

At the same time, in cases outside the federal courts’ diversity orsupplemental jurisdiction (which would probably be most casesagainst State actors), Aleinikoff’s proposal would relegate the enforce-ment of customary international law against the States to the Statesthemselves, which would be free to declare such law completely inap-plicable, as the citizens of Oklahoma have done, and could evenadopt a reverse–Charming Betsy rule. In this respect, Aleinikoff’s pro-posal is open to the same objection as Ramsey’s. Aleinikoff acknowl-edges that some might think that his proposal “renders [customaryinternational law] useless where it is most needed.”466 As notedabove, the Founding generation was most concerned about violationsof international law by the States, yet Aleinikoff’s proposal is designedto leave it to the States to determine the applicability of customaryinternational law to the States (although it would not quite accom-plish this).467

464 See Aleinikoff, supra note 20, at 99. R

465 See id. at 100.466 Id. at 99.467 To be sure, Congress could cure this problem, as well as the “accident of diver-

sity” problem, by conferring exclusive jurisdiction on the federal courts in all casesraising claims under customary international law. As discussed above, such a statutecould be upheld, under current doctrine, even if customary international law werenot regarded as self-executing federal law. See supra text accompanying notes 45–56.In Aleinikoff’s view (unlike Young’s and Ramsey’s) the federal courts would not bebound by the States’ contrary interpretations of such law. Thus, a bare grant of juris-



In response to this objection, Aleinikoff explains that “nothing in[his] understanding of [customary international law] argues againstthe existence of federal common law relating to foreign affairs.”468

And this federal common law of foreign affairs, which would be “bind-ing on [S]tate courts,”469 could include some norms of customaryinternational law. In his view, “[customary international law] and fed-eral common law cases involving foreign affairs are overlapping . . .categories.”470 Aleinikoff does not develop a theory of federal com-mon law to guide the courts in determining which norms of custom-ary international law properly have that status. He tells us that “notall” such norms do,471 but, as noted above, defenders of the modernposition do not contend otherwise. Thus, Aleinikoff’s position mayamount to a theory for making customary international law applicablein the courts even when it lacks the force of preemptive federal lawunder the modern view. If so, then, far from being an intermediateposition, it would be an expansion of the role of customary interna-tional law in the U.S. legal system beyond that entailed by the modernposition.472

If relatively few norms of customary international law wouldemerge with federal common law status under Aleinikoff’s approachto federal common law, then his proposal would be quite similar toYoung’s.473 Both Aleinikoff and Young tell us, in the end, that theapplicability of customary international law turns on State law exceptwhen federal law requires its application. While Young envisions avery restricted role for federal law, Aleinikoff seems open to a broaderfederal role. To the extent that Aleinikoff would base the preemptiveforce of some customary international law on the existence of a “fed-eral common law of foreign relations” having its source in the consti-

diction over such cases would establish a regime much like the modern position. Butthis result could be achieved only through congressional action. The problem underdiscussion arises only when Congress has not acted, as no one doubts that Congresscould enact the modern position by legislation.468 See Aleinikoff, supra note 20, at 99–100. R

469 Id. at 100.470 Id.471 See id. at 97.472 Aleinikoff recognizes that his proposal would go beyond the modern position

in at least one respect: he would permit later-in-time customary international normsto trump federal statutes and treaties. See id. at 100; supra note 459. R

473 Aleinikoff’s proposal would not suffer from the problems discussed above thatresult from Young’s proposal to use choice-of-law rules to determine the applicabilityof customary international law. On the other hand, Young’s proposal does not sufferfrom the Erie problem that afflicts Aleinikoff’s proposal.



tutional structure, his position resembles that of Professors Bellia andClark, which I examine next.

D. The Bellia-Clark Position

In light of the revisionists’ heavy reliance on the work of Profes-sor Bradford Clark, the recent contribution to this debate of Profes-sors Clark and Bellia holds special interest. Clark pioneered the“Supremacy Clause exclusivity” critique of federal common law thatsome revisionists have found compelling.474 On this view, the threecategories of federal law listed in the Supremacy Clause are the solecategories of preemptive federal law envisioned by the Constitution.The argument relies on constitutional structure as well as text. Clark’sclaim is that the process for making the three forms of federal law wasintentionally made difficult in order to protect the States from havingtheir laws displaced. Thus, Article I’s bicameralism and presentmentrequirements for the making of federal statutes serve as a proceduralsafeguard of federalism, as do Article II’s requirement of two-thirdsSenate consent for making treaties and Article V’s onerous require-ments for amending the Constitution. The central role given to theSenate in each situation reflects the federalism purpose underlyingthese requirements, as the Founders regarded the Senate as the bodymost responsive to State concerns. On this view, the trouble with fed-eral common law—understood as “federal rules of decision whosecontent cannot be traced directly by traditional methods of interpreta-tion to federal statutory or constitutional commands”475—is that theprocess for making it evades the Constitution’s onerous and carefullywrought procedures for preempting State law.476 The revisionistsapply this general critique of federal common law to the more specific

474 See Bradford R. Clark, Federal Common Law: A Structural Reinterpretation, 144 U.PA. L. REV. 1245, 1273–74 (1996) (“For the political safeguards of federalism to func-tion effectively, the federal government must adhere closely to the detailed proce-dures required for the adoption of positive federal law—most notably, therequirements of bicameralism and presentment.”). For a more recent elaboration ofthese arguments by Clark, see Clark, supra note 24, at 1328–46. Of the revisionists, RYoung relies on this argument most heavily, see supra text accompanying note 372, but RBradley and Goldsmith invoke it as well, albeit without attributing the argument toClark, see Bradley & Goldsmith, Critique, supra note 9, at 868; see also RAMSEY, supra Rnote 17, at 289-90 (articulating a similar view). For a critique of this view, see RMonaghan, supra note 167. R475 HART & WECHSLER, supra note 55, at 607. R476 Once State law has already been preempted, the carefully wrought procedures

for making federal law will actually make it difficult to devolve power to the States. SeeVazquez, supra note 116. However, in the beginning, the procedures did operate to Rprotect State law from being preempted.



argument that customary international law has the status of federalcommon law.

In their recent contribution to the customary international lawdebate, Bellia and Clark begin by agreeing with the revisionists thatcustomary international law is not, as such, made preemptive of Statelaw by the Supremacy Clause.477 They stress that the frequent state-ments in pre-Independence British sources that the law of nations waspart of the law of the land did not necessarily mean that it was supe-rior to other parts of the law of the land.478 In fact, they point out,parts of the law of nations, such as the law merchant, were understoodto be subject to local variation.479 Thus, the new nation’s inheritanceof that principle from England does not have implications for its placein the hierarchy of domestic law. They also agree with the revisioniststhat Article III does not give the federal courts the power to articulateand enforce norms of customary international law as preemptive fed-eral law.480

Bellia and Clark go on to argue, however, that a subset of the lawof nations was deemed applicable in the courts on a completely differ-ent rationale. Specifically, “history and structure demonstrate thatcourts have applied certain principles derived from the law of nationsas a means of upholding the Constitution’s allocation of foreignaffairs powers to Congress and the President—in particular, the pow-ers to recognize foreign nations and decide questions of war andpeace.”481 The portion of the law of nations that was considered judi-cially enforceable on this theory was the portion that gave rise to “per-fect rights”—that is, rights of sovereigns whose violation was regardedas a just cause for war.482 The sources cited by Bellia and Clark estab-

477 See Bellia & Clark, supra note 21, at 34–37. R

478 See id. at 39.479 See id. at 20.480 See id. at 39.481 Id. at 7.482 See id. at 17. They trace the concept of “perfect rights” to Vattel and other

writers on the law of nations. See JEAN JACQUES BURLAMAQUI, THE PRINCIPLES OF NATU-

RAL AND POLITIC LAW 173 (Thomas Nugent trans., Cambridge, Univ. Press 4th ed.1792) (1747 & 1751); ALBERICO GENTILI, DE IURE BELLI LIBRI TRES 106 (John C. Rolfetrans., Clarendon Press 1933) (1612); G.F. DE MARTENS, SUMMARY OF THE LAW OF

NATIONS, FOUNDED ON THE TREATIES AND CUSTOMS OF THE MODERN NATIONS OF

EUROPE 273 (William Cobbett trans., Phila., Thomas Bradford 1795); SAMUEL

PUFENDORF, DE JURE NATURAE ET GENTIUM LIBRI OCTO 127 (C.H. Oldfather & W.A.Oldfather trans., Clarendon Press 1934) (1688); EMMERICH DE VATTEL, THE LAW OF

NATIONS 145 (Joseph Chitty ed., Philadelphia, T. & J.W. Johnson & Co. 1883) (1758);CHRISTIAN WOLFF, JUS GENTIUM METHODO SCIENTIFICA PERTRACTATUM 43–108 (JosephH. Drake trans., Clarendon Press 1934) (1764).



lish that the Constitution presumptively requires adherence to thesenorms in order to avoid triggering foreign conflict. “The law ofnations supplied a well-known set of default rules relating to perfectrights that would have informed the Founders’ understanding of thefederal political branches’ recognition and war powers.”483

Bellia and Clark acknowledge that “[n]one of [the early] casesexpressly involved preemption of [S]tate law; in each, the federalcourt exercised exclusive jurisdiction, and [S]tate law did not purportto provide a conflicting rule of decision.”484 But their theory clearlyrequires the conclusion that State laws in conflict with the relevantrules of customary international law infringe the federal politicalbranches’ exclusive powers over recognition and war and peace andthus may not be enforced. “Were . . . states free to disregard perfectrights, they could trigger a war and usurp the political branches’exclusive powers over war and peace.”485 The allocation of such pow-ers to the federal government is, on the Bellia-Clark view, a feature ofthe Constitution, and, of course, “[t]he Constitution preempts con-flicting [S]tate law.”486 Thus, Bellia and Clark conclude that, in theseearly cases, “[b]y reasoning that the constitutional allocation of pow-ers required adherence to certain principles of the law of nations, theCourt laid the groundwork for its future holding [in Sabbatino] thatsuch principles apply in both [S]tate and federal court, and even inthe face of contrary [S]tate law.”487

Because the preemptive force of (some) customary internationallaw is based on an inference from the Constitution itself, the Bellia-Clark argument is consistent with Clark’s “Supremacy Clause exclusiv-ity” argument. The “Supremacy Clause exclusivity” objection to fed-eral common law has traction only to the extent that the preemptiveforce of such law cannot be traced to one of the categories of federallaw listed in the Supremacy Clause. But much of what is oftenregarded as (illegitimate) federal common lawmaking can be justifiedas (legitimate) statutory or constitutional interpretation. Clark recog-nized this point in developing his original critique of federal commonlaw,488 and his work with Bellia on customary international law fitscomfortably within this tradition.

Bellia and Clark do not fully work through the implications oftheir historical research for the modern debate about the status of

483 Bellia & Clark, supra note 21, at 46. R484 Id. at 75.485 Id. at 59.486 Id. at 75.487 Id. at 58. Their treatment of Sabbatino is discussed below.488 See Clark, supra note 474, at 1289. R



customary international law. They “do[ ] not attempt to resolve allquestions regarding the status of customary international law in U.S.courts,” and they “take no position on how courts should treat numer-ous modern rules of customary international law unknown to theFounders.”489 But they do not deny that their thesis would supportmany features of that position. In my view, their structural argumentbroadly supports the modern position.

Bellia and Clark argue only that a subset of the law of nations, asknown to the Founders and the early Court, was thought to provide adefault rule against which to measure intrusions on the federal gov-ernment’s exclusive powers of recognition and war. This subset con-sisted of norms that created “perfect rights,” meaning norms whoseviolation justified resort to war by offended nations.490 The fact thatonly a subset of the law of nations was covered might be thoughtinconsistent with the claim—often attributed to defenders of themodern position—that all of modern customary international law hasthe status of preemptive federal law. But, on closer inspection, theBellia-Clark position may be consistent with even that broad claim. AsBellia and Clark explain, the law of nations was thought, in the Foun-ders’ time, to include the law merchant and general commercial law,as well as what Bellia and Clark call the law of state-to-state relations.Although they do not say so clearly, it appears that Bellia and Clarkequate the portion of the law of nations that gave rise to “perfectrights” with the portion of the law of nations that regulated state-to-state relations.491 When Bellia and Clark note that not all of the law ofnations was thought to have the preemptive force they claim for “per-fect rights,” they appear to be excluding only the law merchant andgeneral commercial law—which, as they show, was always thought tobe subject to local variation.492 As already noted, these latter branchesof the old law of nations are today not regarded as public interna-tional law at all; rather, they are considered part of the conflict of laws

489 Bellia & Clark, supra note 21, at 90. R490 See id. at 17491 See id. at 20, 24–26, 38, 41, 42, 56–58. They suggest that the role of the courts

was understood to be only to protect the perfect rights of foreign nations. This possi-ble limitation on the context in which the law of state-to-state relations was regardedas enforceable and preemptive federal law is discussed below.492 See id. at 15, 20–22; see also H.W. HALLECK, ELEMENTS OF INTERNATIONAL LAW

AND LAWS OF WAR 130, 134 (Philadelphia, J.B. Lippincott & Co. 1866) (citing as exam-ples of “imperfect” duties “the ordinary duties of comity [and] of diplomatic andcommercial intercourse”). By “diplomatic . . . intercourse,” Halleck means “those ofsending and receiving diplomatic legation,” which are “at most imperfect obligations.”See id. at 134. Today, there is no international law duty to send or receive a diplo-matic legation.



or merely domestic law. Defenders of the modern position do notclaim that these norms have the status of preemptive federal law.493

The precedents discussed by Bellia and Clark might be thoughtirrelevant to the modern debate for a different reason. Their argu-ment regarding the preemptive force of norms conferring perfectrights appears to depend on the notion that, under the law of nationsas it then prevailed, violation of public rights was a just cause for initi-ating war. This is, indeed, the rationale for regarding violations ofsuch norms to be an intrusion into the political branches’ exclusivewar powers. Today, however, international law forbids the use offorce to resolve international disputes. The use of force is permittedonly in self-defense or if authorized by the Security Council (in whichthe United States holds a veto).494 It is thus no longer the case thatthe United States’ breaches of other nations’ rights under interna-tional law will justify the use of force against the United States by thosenations. Moreover, it might be argued that, given the United States’superior military power, the use of force by other nations against theUnited States is, in any event, highly unlikely today. Thus, to theextent that the historical argument rests on the allocation of the warpower to Congress and the idea that breaches of perfect rights justi-fied war, it may be thought to be irrelevant to modern conditions.

But Bellia and Clark do not rely solely on the federal govern-ment’s excusive power over war and peace. They also invoke theexclusive recognition power, which is itself an inference from the allo-cation to the President of the power to send and receive ambassa-dors.495 Since territorial sovereignty is an incident of statesovereignty, any state action that interfered with such sovereigntywould be in conflict with the federal government’s decision to recog-nize a foreign state as a state in the international sense. More broadly,Bellia and Clark write:

[T]he Constitution assigns the power to send and receive ambassa-dors to the President and the Senate. This power cannot be under-

493 See also infra text accompanying notes 495–96 (noting that, to the extent that RBellia and Clark rely on the federal government’s exclusive recognition power, theirargument would seem to extend to the state-to-state branch of the law of nations in itsentirety).494 See U.N. Charter art. 2, para. 4; id. art. 51; cf. id. art. 27 (regarding the Security

Council veto).495 See RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW § 204 cmt. a (1987) (stat-

ing that the President’s recognition power “is implied in the President’s express con-stitutional power to appoint Ambassadors (Article II, Section 2) and to receiveAmbassadors (Article II, Section 3), and his implied power to conduct the foreignrelations of the United States”).



stood without reference to what it meant to recognize anothernation at the founding. Recognition implied that the United Stateswould respect the rights of the state in question under the law ofnations.496

On this theory, the act of recognizing a foreign state implicitly incor-porates the state’s rights vis-a-vis the United States under internationallaw, not just those whose violation justifies the use of force. If so, thenall violations of international legal rights of recognized foreign stateswould be preempted, not just those that justified war.

Additionally, it might be argued that violations of other states’rights under international law are preempted because they threatenforms of retaliation not involving the use of military force, to the ulti-mate detriment of the nation as a whole. There is an obvious struc-tural reason to deny a single State the power to impose these costs onthe nation as whole, as the Founders recognized. Even if such viola-tions do not threaten military invasion, they produce internationalfriction, which, in turn, makes it more difficult for the federal govern-ment to accomplish its foreign relations objectives through the negoti-ation of treaties or simply through diplomacy. This argument forregarding State laws violating customary international law as pre-empted by the constitutional structure is less directly linked to anexclusive federal power expressly in the Constitution’s text than the“perfect rights” argument is to the war power. But the Constitution’sallocation to the federal government of the power to send and receiveambassadors and the power to make treaties might be thought toimply an exclusive federal power to determine whether our relationswith other nations shall be friendly or hostile. This may be what Belliaand Clark have in mind when they refer to the federal government’sexclusive foreign affairs power.497

The argument that State violations of the state-to-state branch ofthe law of nations are preempted because they interfere with the fed-eral government’s exclusive power over foreign relations is essentiallythe structural argument for the modern position discussed in Part I.So understood, the allocation-of-powers rationale for regarding Stateviolations of customary international law to be preempted might bethought to prove too much, as it would suggest that all State acts thatare unfriendly to foreign nations or would otherwise produce interna-tional friction are preempted. Indeed, the Bellia-Clark argument hasobvious affinities to the much-maligned dormant foreign affairs ratio-

496 Bellia & Clark, supra note 21, at 89. R

497 See id. at 28–29.



nale of Zschernig v. Miller.498 In that case, the Supreme Court foundan Oregon statute to be invalid as an unconstitutional interferencewith the federal government’s excusive power to conduct foreign rela-tions. What exactly about the State law made it invalid on this groundthe Court did not make entirely clear, and Zschernig has been criti-cized as overbroad by many of the same scholars who have advancedrevisionist views of the status of customary international law.499 Cer-tainly, the broad claim that all State conduct that has an effect onforeign relations is preempted would invalidate too much State con-duct that has traditionally been regarded as proper. In today’s global-ized world, almost everything that States do has at least an incidentaleffect on foreign relations.500 Zschernig remains the only case to find aState law invalid on dormant foreign relations grounds and scholarshave suggested that it would not be followed today.501

On the other hand, Zschernig has not been overruled. I haveargued elsewhere that the decision’s basic insight that some Stateaction is invalid because it unduly interferes with the federal govern-ment’s exclusive power to conduct foreign relations is sound.502 Thechallenge is to transform that insight into workable doctrine. In Amer-ican Insurance Ass’n v. Garamendi,503 the majority and dissenting opin-ions offered competing limiting constructions of the decision. JusticeSouter, for the majority, proposed that Zschernig be read to establish arule of “field preemption”504 for State conduct outside areas of tradi-tional State regulation, but a rule of “conflict preemption” for Stateconduct within such areas.505 Justice Ginsburg, for the dissenters, pro-posed limiting Zschernig to State laws that “reflect[ ] a State policy criti-cal of foreign governments and involve[ ] sitting in judgment on

498 389 U.S. 429 (1968). Bellia and Clark acknowledge the relevance of Zschernigin a footnote. See Bellia & Clark, supra note 21, at 91 n.489. R

499 See Bradley & Goldsmith, Critique, supra note 9, at 865; Jack L. Goldsmith, Fed- Reral Courts, Foreign Affairs, and Federalism, 83 VA. L. REV. 1617, 1631 (1997); Ramsey,supra note 17, at 563–70; Young, supra note 444, at 36–37; Michael D. Ramsey, The RPower of the States in Foreign Affairs: The Original Understanding of Foreign Policy Federalism,75 NOTRE DAME L. REV. 341, 357–65 (1999).500 Young makes this point forcefully. See Young, supra note 19, at 420–23. R

501 See Bradley & Goldsmith, Critique, supra note 9, at 865; Ramsey, supra note 17, Rat 563. But cf. Bellia & Clark, supra note 21, at 91 n.489 (“[T]he rationale for constitu- Rtional preemption in Zschernig is an extension of the rationale for preemption inSabbatino.”).502 See Vazquez, W(h)ither Zschernig?, supra note 86, at 1268, 1318. R

503 539 U.S. 396 (2003).504 See id. at 419 n.11.505 See id. at 419.



them.”506 I have proposed a construction similar to Ginsburg’s, underwhich State laws that single out foreign states or groups of states foradverse treatment would be preempted.507 Any of these possible ruleswould be workable doctrinal manifestations of Zschernig’s insight thatthe Constitution’s structure—specifically, its allocation of the powerto conduct foreign relations exclusively to the federal government—implicitly preempts State laws or conduct that unduly interfere withsuch relations. The view that State law or conduct that violates cus-tomary international law is similarly preempted might be regarded asan additional doctrinal manifestation of this structural principle—amanifestation that, as Bellia and Clark have demonstrated, has astrong historical pedigree.

Bellia and Clark clearly do not regard the structural argument foraccording preemptive force to some customary international law to bedependent on the allocation of the war power to the federal govern-ment, or the fact that international law once authorized the use offorce in response to violations, as they argue that the Supreme Court’sact-of-state decision in Sabbatino rests on this very structural argument.By the time Sabbatino was decided, international law no longer counte-nanced the use of force in response to breaches of international lawby other nations.508 Noting that “[r]ules like the act of state doctrinehave deep roots in the traditional perfect right of territorial sover-eignty under the law of nations,”509 Bellia and Clark argue that Sabba-tino “is best understood overall as a consequence of the Constitution’sallocation of foreign affairs powers to the political branches of thefederal government.”510 Indeed, they regard Sabbatino as makingexplicit for the first time what was implicit in the earlier cases: thatcustomary norms giving rise to perfect rights of foreign states preemptState law.511

As a manifestation of the Bellia-Clark allocation-of-powers ratio-nale for according preemptive force to (some) customary interna-tional law, Sabbatino poses serious challenges. The Court did holdthat the act-of-state doctrine has its basis in domestic separation-of-powers principles, and it also held that this doctrine was binding onthe States. But, as Bellia and Clark note, the Court recognized thatthe act-of-state doctrine was not itself a principle of international

506 See id. at 439 (Ginsburg, J., dissenting) (quoting HENKIN, supra note 230, at R164) (internal quotation marks omitted).507 See Vazquez, W(h)ither Zschernig?, supra note 86, at 1321. R508 The U.N. Charter came into force in 1945. Sabbatino was decided in 1964.509 Bellia & Clark, supra note 21, at 85 n.459. R510 Id. at 84.511 See id. at 87.



law.512 Sovereign states did not have a perfect right to have foreigncourts give unquestioning effect to all of their public acts performedwithin their territory. Thus, as Bellia and Clark note, the Court inSabbatino “gave foreign sovereigns greater protection than they enjoyedunder customary international law at the time.”513 In this light, Sabba-tino might be viewed as an allocation-of-powers decision having someparallels to the principle that Bellia and Clark identify (namely, thatthe constitutional structure prohibits State violation of customaryinternational law conferring perfect rights on foreign sovereigns), butnot as a manifestation of that principle. Indeed, Sabbatino wouldappear to be in tension with the latter principle insofar as the Courtheld that the public acts of foreign states within their own territorymay not be challenged in certain circumstances even if they violate cus-tomary international law.

Bellia and Clark claim that Sabbatino is a manifestation of the ideathat, because of its allocation of the foreign relations power to thefederal government, the Constitution preempts state law that conflictswith certain norms of customary international law but not others.Specifically, they argue that the Supreme Court held that “the Consti-tution required federal and state courts both to adhere to a rule basedon traditional notions of territorial sovereignty (the act of state doc-trine), and to disregard a more recent rule of customary internationallaw curtailing territorial sovereignty (the proffered prohibition againstdiscriminatory, uncompensated takings).”514 But this reading of thecase fails to establish that the Court was requiring compliance with thelaw of nations. After all, the second, more recent norm of interna-tional law (barring discriminatory or uncompensated takings) pur-ports to be a limitation of the first, older norm of international law(protecting the state’s territorial sovereignty). Together, they tell usthat the state’s sovereignty over its own territory does not include theright to take the property of foreign nationals in a discriminatory fash-ion or without providing compensation. If the Court was insisting thatfederal and State courts give effect to the public acts of foreign stateseven when international law does not give the States a right to per-form such acts, then it was not according preemptive force to anyexisting rule of international law.

Bellia and Clark seek to avoid this conclusion by arguing that Sab-batino “highlights an important point about the interaction of perfect

512 See Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398, 421 (1964).513 Bellia & Clark, supra note 21, at 86. R

514 Id. at 85–86 (footnote omitted).



rights and the Constitution over time.”515 In their view, the act-of-state doctrine gives preemptive force to some norms of internationallaw conferring rights on foreign states, unmodified by more recentnorms of international law that cut back on such rights. They write:

The Court has been willing to add to, but not subtract from, the listof sovereignty-respecting default rules that courts are required toapply in order to uphold the Constitution’s allocation of powers. InThe Schooner Exchange and The Paquette [sic] Habana, the Court aug-mented the traditional list of perfect rights with later-emerging cus-toms favoring foreign sovereigns—namely, the immunity of foreignwarships and fishing boats. Judicial adherence to these customsupheld the Constitution’s allocation of powers in much the sameway as judicial adherence to the original perfect rights of sovereigns.In both cases, the rule in question avoided judicial initiation or esca-lation of hostilities, and thus preserved political branch control overwar and peace. In Sabbatino, however, the Court refused to departfrom a traditional rule of territorial sovereignty (historicallyregarded as a perfect right), even though the Court acknowledgedthat the community of nations no longer recognized absolute terri-torial sovereignty. The Court based this refusal on the Constitu-tion’s allocation of powers as well. In effect, the Court held that anydecision to abandon the traditional perfect rights of recognized for-eign sovereigns would foster resentment and thus should be madeby the political branches rather than the courts or the states.516

Bellia and Clark here seem to envision a one-way ratchet. New rulesof international law that expand the rights of foreign sovereigns areautomatically added to the list of international law norms that pre-empt inconsistent State law, but new rules that restrict such rights donot preempt State law unless incorporated by federal statute ortreaty.517

This rule may or may not be appealing from a structural constitu-tional standpoint, but it does not capture the Court’s holding or rea-

515 Id. at 88.516 Id. at 88–89.517 Note that, under this theory, the preemptive force of any given norm depends

not just on the nature of the norm, but on whom one is seeking to apply it against. Achange in international law that expands the rights of foreign states vis-a-vis theUnited States correspondingly restricts the rights of the United States vis-a-vis foreignstates, and vice-versa. Thus, presumably, a norm that restricts territorial sovereigntywould be enforceable against the United States (and the States), while a norm thatexpands the rights of states would be enforceable only insofar as it favors foreignstates. This means that a foreign state’s right not to have its citizen’s property takenby the United States without compensation would be enforceable, whereas the UnitedStates’ right not to have its citizen’s property taken without compensation would notbe.



soning in Sabbatino. First, as already noted, it is difficult to readSabbatino as resting on the preemptive force of any rule of interna-tional law. The Court itself denied that international law requiredstates to give effect to the acts of other states performed within theirown territory.518 The Court noted that a state’s courts are free todecline to give effect to the penal laws of other states.519 Moreover, itis well accepted that the courts of one state may refuse to give effect tothe act of another state performed within its own territory if, for exam-ple, that act contravenes the forum state’s public policy.520 Sincestates are free not to give effect to the public acts of other states per-formed within their own territory even if those acts did not violateinternational law, it cannot be said that international law gives states aright to have their public acts enforced in other states’ courts.

Second, the Court in Sabbatino did not hold that all new rules ofinternational law that restrict foreign states’ preexisting rights of terri-torial sovereignty were inapplicable in our courts. It was careful tolimit its holding to one particular rule of international law: the oneprohibiting the taking of a foreign national’s property in a discrimina-tory fashion or without compensation. It held that norm not to beenforceable in the particular context before it because the normtouched “sharply on national nerves”521 and was at the time the sub-ject of considerable controversy in the international legal community,with many capital-importing states denying that it had the force ofcustomary law.522 The Court left open the possibility that a norm lack-ing these characteristics could be applied by State and federal courtsto invalidate an act of a foreign state within its own territory. Thus,the norm’s inapplicability did not turn on its age but on other factors.

518 See Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398, 421 (1964).519 Id. at 413–14.520 Matusevitch v. Telnikoff, 877 F. Supp. 1, 4 (D.D.C. 1995) (“[L]ibel standards

that are contrary to U.S. libel standards would be repugnant to the public policies ofthe State of Maryland and the United States. Therefore, . . . this court declines torecognize the foreign judgment.”); Bachchan v. India Abroad Publ’ns Inc., 585N.Y.S.2d 661, 665 (Sup. Ct. 1992) (“The protection to free speech and the pressembodied in that amendment would be seriously jeopardized by the entry of foreignlibel judgments granted pursuant to standards deemed appropriate in England butconsidered antithetical to the protections afforded the press by the U.S. Constitu-tion.”); PETER HAY, PATRICK J. BORCHERS & SYMEON C. SYMEONIDES, CONFLICT OF LAWS

§ 24.44 (5th ed. 2010) (“[T]he public policy defense serves as an umbrella for a vari-ety of concerns in international practice which may lead to a denial of recognition [ofa foreign judgment]”).521 See Sabbatino, 376 U.S. at 428.522 See id. at 434–35.



As discussed in Part I, Sabbatino is best read as holding that normsof international law that do not meet certain standards of clarity orbreadth of acceptance lack the force of preemptive federal law insofaras they are sought to be applied to certain acts of foreign sovereigns.The norm that had been invoked to invalidate Cuba’s act in Sabbatinofell in this category and hence, as a matter of federal law, could not beapplied to invalidate Cuba’s act. On this view of Sabbatino, a norm ofinternational law purporting to limit a foreign state’s territorial sover-eignty would be cognizable in U.S. courts if it met the Court’s stan-dard of clarity and breadth of acceptance, and both federal and Statecourts would be either permitted or required to apply it to invalidatethe foreign state’s act.

This seems to me to be the correct reading of Sabbatino, but onlyhalf of it is supported by the allocation-of-powers rationale for accord-ing preemptive force to customary international law. The require-ment that foreign acts of state be given effect even if they clash withnoncognizable norms of international law is supported by the Bellia-Clark allocation-of-powers principle523; the (possible) permissibility ofinvalidating foreign acts of state that clash with cognizable norms ofinternational law is not. It is hard to imagine that the politicalbranches’ ability to conduct foreign relations will be significantly ham-pered by the courts’ refusal to invalidate a foreign act on thisground.524 This part of the Court’s analysis appears to be groundedinstead on countervailing concerns about justice or the rule of law.But the Court’s analysis here is compatible with the allocation-of-powersrationale that drives the rest of its analysis, the idea being that thepolitical branches’ ability to conduct foreign relations will not beunduly undermined by a failure to give effect to foreign state acts thatviolate clear and broadly accepted norms of customary internationallaw, just as it is not unduly undermined by failure to give effect toforeign state acts performed outside their territory. Consequently, the

523 As discussed above, though, Sabbatino does not hold that any particular normof international law has the force of preemptive federal law. See supra text accompa-nying notes 514–22. R

524 This would be true, at least, where international law does not prohibit theUnited States from acquiescing in another state’s violation of international law. Fordiscussion of exceptions to this general rule, see Monica Hakimi, State BystanderResponsibility, 21 EUR. J. INT’L L. 341, 342–44 (2010). For a proposal that internationallaw more broadly prohibit third states from actively supporting another state’s viola-tions of international law, see Lea Brilmayer & Isaias Yemane Tesfalidet, Third StateObligations and the Enforcement of International Law, 43 N.Y.U. J. INT’L L. & POL. (forth-coming 2011) (on file with author).



courts are free to apply such norms to advance the norms’ substantivegoals.

But the allocation-of-powers rationale would support the conclu-sion that norms of customary international law—even newly devel-oped ones, and possibly even disputed ones—are applicable andpreemptive insofar as they restrict the freedom of action of the UnitedStates or the States. Such violations do risk international friction andhence would complicate the federal government’s ability to achieveforeign relations goals. If the preemptive force of these norms stemsfrom the allocation of the foreign relations power to the federal gov-ernment, then clearly these norms can be rendered inapplicable bystatute or treaty. The allocation-of-powers rationale might support,but does not necessarily require, the conclusion that the norm is alsoinapplicable if it clashes with an act of the Executive. Bellia and Clarkleave this question for another day.525 As discussed above, there isdoctrinal support for some presidential power to override customaryinternational law. But the allocation-of-powers argument tells us that,in the absence of authorization from the relevant combinations of thepolitical branches, all State officials are bound.

E. The Bradley-Goldsmith-Moore Position

It might seem odd to include Bradley and Goldsmith amongthose proposing an intermediate position, as they spearheaded therevisionist movement. But their post-Sosa contribution to the debate,with coauthor David Moore, builds on the suggestion in their originalcritique of the modern position that it might be proper for the courtsto apply customary international law as federal law even without an actof federal lawmaking (of the sort envisioned by the SupremacyClause) that purports to accord it such status. Responding to scholarswho claimed that Sosa embraced the modern position, they contendthat Sosa refutes the claim “that all of [customary international law],‘whatever [it] requires,’ is automatically incorporated wholesale intopost-Erie federal common law.”526 But that, of course, leaves open thepossibility that some customary international law has the status of fed-eral law in the absence of incorporation by federal statute or treaty.Their post-Sosa article elaborates on the circumstances in which thismight be the case.

525 See Bellia & Clark, supra note 21, at 75. R526 See Bradley, Goldsmith & Moore, supra note 33, at 902 (second alteration in R

original) (quoting Brilmayer, supra note 88, at 324). As discussed in Part I, Sosa does Rnot refute that proposition. See supra text accompanying notes 234–72. R



This subpart considers the circumstances in which Bradley, Gold-smith, and Moore would accept political branch authorization shortof federal statute or treaty to treat customary international law as fed-eral law. Although they insist that the circumstances in which this isproper are few and limited, their rationale for accepting these tech-niques would appear to extend more broadly.

Some of their examples of permissible applications of customaryinternational law fall in the category of express or implicit incorpora-tion by statute or treaty. For example, they rightly note that all trea-ties implicitly incorporate the norms of customary international lawregarding treaty interpretation, as reflected in Articles 31 and 32 ofthe Vienna Convention on the Law of Treaties.527 Although theseexamples are consistent with the revisionist position, others are in ten-sion with it.

First, they recognize that the statute conferring federal jurisdic-tion over interstate disputes can properly be read to authorize theelaboration of federal common law drawn from customary interna-tional law. In the case of interstate disputes, there are “uniquely fed-eral interests derived from the Constitution [that] demand a federalrule,”528 and, in crafting these rules, the Court properly draws uponcustomary international law because these disputes “are directly analo-gous to disputes between nations.”529 In distinguishing this rationalefor treating some customary international law as federal law from themodern position, Bradley, Goldsmith, and Moore stress that, evenbefore Erie, the interstate jurisdiction clause was understood toauthorize federal courts to apply federal judge-made rules without leg-islative authorization, whereas “[f]or over 200 years, courts have notperceived a structural need to apply [customary international law] asfederal common law” more generally.530 They argue that the absenceof a “long historical pedigree” renders the claim of structural necessityfor regarding customary international law as federal law less plausible

527 Vienna Convention on the Law of Treaties arts. 31–32, opened for signature May23, 1969, 1155 U.N.T.S. 331. As Bradley, Goldsmith, and Moore note, this treaty isbinding on the United States to the extent it reflects customary law, as the UnitedStates is not a party to it. Bradley, Goldsmith & Moore, supra note 33, at 921–22. RBradley, Goldsmith, and Moore seem to doubt the propriety of applying the custom-ary international law of treaties to determine whether a treaty exists in the first place.To be sure, it is a stretch to say that the law concerning the existence of a treaty isrelevant because it is incorporated into the treaty, when the very question is whetherthere even is a treaty. On the other hand, such law could easily be understood to beincorporated into the Constitution’s provisions addressing the making of treaties.528 Bradley, Goldsmith & Moore, supra note 33, at 915. R529 See id. at 916.530 See id. at 917.



in the latter case.531 This argument overlooks the lengthy periodbefore Erie in which the federal courts had jurisdiction in virtually allcases implicating customary international law and enforced their ownbest understanding of what that law required. Although the Stateswere understood to have the power to depart from general commer-cial law, it was not clear that they were understood to have the powerto override the state-to-state portion of the law of nations, and revi-sionists have not identified examples of the States doing so.532 Giventhese features of the judicial treatment of customary international lawbefore Erie, the “structural necessity” of treating customary interna-tional law as federal law as opposed to general law was not apparentuntil Erie interred the general law. Indeed, the long history of treat-ing customary international law as something very different fromtoday’s State law provides its own lengthy pedigree for its status asfederal law.

Bradley, Goldsmith, and Moore also accept the federal laws con-ferring federal jurisdiction over admiralty cases as a source of author-ity for the articulation of a federal law of admiralty, “a traditionalcomponent of the law of nations that was important to the prosperityof the infant United States.”533 Here, too, they distinguish thebroader claim that customary international law should be treated asfederal law, noting that the Constitution’s express reference to admi-ralty is “in notable contrast with the treatment of the law of nationsmore generally.”534 But the contrast is not very sharp if one considersthat the Constitution’s reference to admiralty comes in a jurisdictionalprovision and that the Founders also conferred jurisdiction on thefederal courts over the categories of cases most likely to implicate thelaw of nations as a means of advancing their “well-documented desireto ensure that [S]tates complied with international law.”535 The grantof federal jurisdiction could be expected to achieve this goal giventheir understanding that the law of nations would be judiciallyenforceable, at a minimum, as general law regardless of State or fed-eral incorporation. Bradley, Goldsmith, and Moore are willing to rec-ognize that admiralty law has the status of federal law today eventhough, as they note, it was clearly held not to be federal law beforeErie.536 For the reasons discussed by Bellia and Clark, the structural

531 See id. at 918.532 See supra text accompanying notes 479–87, 491. R

533 See Bradley, Goldsmith & Moore, supra note 33, at 918. R

534 See id.535 See id. at 883.536 See id. at 892.



and historical case for treating the state-to-state branch of the law ofnations as federal law is, if anything, stronger.537

Finally, and perhaps most illuminatingly, Bradley, Goldsmith, andMoore are willing to accept the federal status of some norms of cus-tomary international law if “incorporated” into U.S. law by the execu-tive branch.538 As an example, they offer the norms regardingimmunity for foreign states and heads of state. Before Erie, the courtsapplied these norms of customary international law as general lawwithout authorization. At around the time that Erie was decided, theSupreme Court indicated that the courts were to regard case-specificexecutive branch statements regarding immunity as binding, and, incases in which the executive branch had not expressed an opinion,the courts were to apply the principles of customary international lawon immunity that the executive branch had accepted.539 In 1976,Congress codified the international law concerning the immunity offoreign states in the FSIA, but the Act is unclear regarding its applica-bility to heads of state or foreign officials generally.540 According toBradley, Goldsmith, and Moore, of the courts that read the FSIA notto cover the immunity of heads of state, “some recognize [the] immu-nity only in the face of an explicit suggestion of immunity by the Exec-utive,” whereas “[o]thers rely on the lack of an executive branchsuggestion simply as a factor weighing against immunity.”541 All, how-ever, look for executive branch authorization “at least to somedegree.”542

It is surprising that scholars who object to free-wheeling federalcommon lawmaking would be so receptive to the incorporation of cus-tomary international law into federal law by the executive branch act-ing alone. After all, the main objection to federal commonlawmaking is that it evades the carefully wrought procedures set forth

537 See also Bederman, supra note 64, at 348 (“To suggest . . . that originalist Rgrounds for exceptionalism in admiralty as ‘law of the land’ are stronger than theconstitutional basis for disputes arising under the law of nations seems wholly extrava-gant. Put another way, there seems to be no sensible originalist argument for accord-ing [the general maritime law] greater jurisprudential status than [customaryinternational law].”).538 See id. at 935–36.539 See id. at 922–23.540 See id. The Court has since held that the FSIA does not address the immunity

of foreign state officials. See Samantar v. Yousuf, 130 S. Ct. 2278 (2010).541 Id. at 924.542 Id.



in the Constitution for making preemptive federal law.543 Asexplained above, the claim is that the hurdles that must be overcometo make federal law were put in place to protect the States from hav-ing their laws displaced too easily. From this perspective, “lawmaking”by the executive branch acting alone should be similarly problematic,as it does not satisfy the bicameralism requirement for federal statutesor the requirement of Senate consent to treaties.544 To attribute thepreemptive force of these norms to ad hoc executive branch decision-making seems more offensive to rule-of-law values than a more gener-ally applicable structural argument for preemption, as a retailexecutive incorporation regime would transform the federal courtsinto the federal Executive’s errand boys.545

Executive branch incorporation of customary international lawwith respect to the subset of customary international law that relatesto the immunity of foreign states and officials (including heads ofstate) might perhaps be defended as an exercise of the President’sconstitutional authority to send and receive ambassadors, which hasbeen construed to give the President the power to recognize foreignstates and governments as well. This recognition power has beenacknowledged as the source of the President’s power to enter intoexecutive agreements that have the force of preemptive federal lawwithout the consent of the Senate.546 But revisionists have criticizedthese decisions, too, as incompatible with the aspects of the constitu-

543 See Clark, supra note 474; Thomas W. Merrill, The Common Law Powers of Federal RCourts, 52 U. CHI. L. REV. 1 (1985); Thomas W. Merrill, The Judicial Prerogative, 12 PACE

L. REV. 327 (1992).544 See Bradford R. Clark, Domesticating Sole Executive Agreements, 93 VA. L. REV.

1573, 1597–1607 (2007).545 Justice Douglas used the “errand boy” language in his concurrence in First

National City Bank v. Banco Nacional de Cuba, 406 U.S. 759, 773 (1972) (Douglas, J.,concurring in the result). A majority of the Court in that case rejected the Bernsteinexception, under which the applicability of the act-of-state doctrine would turn on theExecutive’s case-by-case determination that adjudicating the case under internationallaw would not harm the nation’s foreign relations interests. See id. at 772–73; id. at773 (Powell, J., concurring in the judgment); id. at 776–77 (Brennan, J., dissenting).The Bernstein exception takes its name from Bernstein v. N.V. Nederlandsche-Amerikaans-che Stoomvaart-Maatschappij, 173 F.2d 71 (2d Cir. 1949), mandate amended by 210 F.2d375 (2d Cir. 1954), and Bernstein v. Van Heyghen Freres Societe Anonyme, 163 F.2d 246(2d Cir. 1947).546 See United States v. Pink, 315 U.S. 203, 229 (1942); United States v. Belmont,

301 U.S. 324, 330 (1937).



tional structure that they assert are offended by according federal sta-tus to customary international law.547

An alternative explanation of the immunity decisions would be toread them as assuming that customary international law has the forceof federal law for structural reasons, but that the courts are to hewclosely to the executive branch’s positions out of deference to its supe-rior expertise and responsibilities in this area. The proper scope andcontexts of judicial deference to the Executive on questions of cus-tomary international law are beyond the scope of this Article. TheCourt does not accord nearly as much deference to the Executive inother contexts. For example, in the context of the act-of-state doc-trine, the Court has refused to treat the views of the Executive regard-ing the content of international law as binding.548 But perhaps amore deferential approach is warranted for the immunity issuebecause it bears upon jurisdiction rather than the merits, or becauseimmunity is more closely tied to the President’s recognition power.549

The recent Samantar decision may suggest that the Court favorsan executive incorporation approach with respect to immunity issues.As discussed above, the executive branch there argued that foreignofficials were entitled to a “common law” immunity whose contourswere to be determined by reference to “[p]rinciples adopted by theExecutive Branch, informed by customary international law.”550 TheCourt agreed that the immunity of foreign officials was “governed bythe common law.”551 Although it did not discuss the content or statusof this immunity, it did describe the pre-FSIA regime as one in whichthe courts followed the executive branch’s case-specific suggestions ofimmunity and, if the executive branch declined to express its views onthe case, “a district court inquired ‘whether the ground of immunity isone which it is the established policy of the [State Department] to

547 See Young, supra note 19, at 415–17, 432; see also Clark, supra note 544, at R1647–48 (arguing that the Court’s modern reliance on Belmont and Pink to supportan expansive view of sole executive agreements is mistaken).548 See supra note 214. R549 The greater deference that the courts gave to the executive branch on ques-

tions of immunity in the pre-FSIA era may also have stemmed from the unsettlednature of the relevant international law at the time, which was moving from a rule ofabsolute immunity to one of restrictive immunity. See Yelin, supra note 85. As dis- Rcussed below, it is appropriate for the courts to follow the executive branch’s viewswith respect to norms of international law that are unsettled or evolving, as the execu-tive branch speaks for the nation on the international plane regarding how such lawshould develop. See infra text following note 608. R550 See Brief for the United States as Amicus Curiae Supporting Affirmance, supra

note 196, at 8. R551 See Samantar v. Yousuf, 130 S. Ct. 2278, 2284 (2010).



recognize.’”552 The prominent role given to the executive branch inthe pre-FSIA regime, as the Court described it, may suggest that thecourts apply customary international law in this context as filteredthrough the executive branch. If so, the question would remainwhether the contemplated filtering role is peculiar to the immunitycontext, given its close link to the recognition power, or is morebroadly applicable. Very likely, the Court did not mean to express anyview about the nature of the executive branch’s role regarding theimmunity enjoyed by foreign officials,553 much less about its roleregarding other areas of customary international law. Since the issuebefore them was merely whether the FSIA applied, and since the Jus-tices who joined or concurred in Justice Stevens’s opinion had previ-ously expressed varying views concerning the domestic status ofcustomary international law,554 the opinion is best read to leave all ofthese questions open.555

552 Id. (alteration in original) (quoting Republic of Mexico v. Hoffman, 324 U.S.30, 36 (1945)).553 But cf. Statement of Interest of the United States, supra note 197, at 6 (citing R

Samantar for the proposition that, when the State Department has taken a positionregarding the immunity of a foreign official in the particular case, “the court shouldaccept and defer to the determination that Defendant is not immune from suit”).554 See Sosa v. Alvarez-Machain, 542 U.S. 692, 697 (2004) (Souter, J., delivering the

opinion of the Court, with Part IV joined by Stevens, O’Connor, Kennedy, Ginsburg,and Breyer, JJ.); id. at 739 (Scalia, J., concurring in part with Rehnquist, C.J., andThomas, J., joining).555 It might perhaps be argued that the Court’s references to a “common law” of

immunity rather than a customary international law of immunity shows that it was notcontemplating the judicial enforcement of customary international law at all, even asfiltered through the executive branch, but was instead recognizing an executivebranch power to make law, untethered to customary international law, based on thePresident’s authority to conduct foreign relations. This reading seems implausible.The executive branch does not have the power to immunize a private party from suit,no matter how greatly it would advance the nation’s foreign relations. Presumably,the executive branch is, at the very least, required to defend its suggestion of immu-nity by reference to international law as it exists or as the executive branch would likeit to evolve. And, in fact, the State Department does explain its suggestions of immu-nity by reference to international law. See Letter from Harold Hongju Koh, LegalAdviser to the Dep’t of State, to Tony West, Assistant Att’y Gen., Re: Yousuf v.Samantar, Civil Action No 01-13760 (E.D. Va.) (Feb. 11, 2011), reprinted in Statementof Interest of the United States of America, supra note 197, exhibit 1. If a link to Rinternational law is required, then it would appear that the common law that thecourts are enforcing is (at least) customary international law as filtered through theexecutive branch. Indeed, the opinion in Samantar appears to use the term “commonlaw” as a synonym for customary international law, perhaps reflecting the “settled”view that customary international law is federal common law. See Samantar, 130 S. Ct.at 2289 (“[O]ne of the primary purposes of the FSIA was to codify the restrictivetheory of sovereign immunity, which Congress recognized as consistent with extant



In any event, the executive incorporation regime, as applied bythe Supreme Court in pre-FSIA immunity cases, is closer to the mod-ern position than to the revisionist one. Adherents of the modernposition recognize that the executive branch speaks for the UnitedStates at the international plane regarding the content of customaryinternational law and that the courts generally defer to the executivebranch when interpreting such law, especially when the law is in flux.The executive incorporation thesis thus appears to differ from themodern position only insofar as it would altogether disable the courtsfrom disagreeing with the executive branch’s position regarding thecontent of customary international law.556 The difference appears tobe between absolute and significant deference to the Executive. Anacross-the-board rule of absolute deference would appear to conflictwith the Supreme Court’s approach to the application of customaryinternational law in other contexts and, thus, presumably does notapply outside the context of immunity for foreign officials (if eventhen).557 But the more significant point, for present purposes, is thatthe two approaches are far closer to each other than to the approachthat revisionists had attributed to Judge Hand with approval, underwhich the issue would be governed by State law.

* * * *

In sum, the intermediate positions are problematic to the extentthat they leave the question of the applicability and interpretation ofcustomary international law to the States, which they would all do tosome extent. Ramsey’s position is not intermediate at all, as he wouldapparently regard customary international law as always nonpreemp-tive unless incorporated by statute or treaty, whereas Bradley andGoldsmith’s original critique left room for the application of such law

international law. We have observed that a related purpose was ‘codification of interna-tional law at the time of the FSIA’s enactment,’ and have examined the relevant com-mon law and international practice when interpreting the Act. Because of thisrelationship between the Act and the common law that it codified, petitioner argues thatwe should construe the FSIA consistently with the common law regarding individualimmunity, which—in petitioner’s view—was coextensive with the law of state immu-nity and always immunized a foreign official for acts taken on behalf of the foreignstate.” (emphasis added) (citations omitted) (quoting Permanent Mission of India toUnited Nations v. City of N.Y., 551 U.S. 193, 199 (2007))).556 Perhaps the thesis would also disable the courts from applying a norm of cus-

tomary international law on which the executive branch has not yet taken a position,but this situation is highly unlikely to arise.557 For example, in Sabbatino, the Court preferred to declare customary interna-

tional law inapplicable to a whole category of cases than to adopt a rule of absolutedeference. See supra note 214. R



as federal law if incorporated by the executive branch. Young,Aleinikoff, and Bradley, Goldsmith, and Moore would, to varyingdegrees allow customary international law norms not incorporated bytreaty or statute to be treated as federal law in certain circumstances,but (except for Aleinikoff, who is not clear on this point) wouldaccord it such force only in very limited circumstances. In addition,the specific proposals of Ramsey, Young, and Aleinikoff pose signifi-cant problems of application. Bellia and Clark offer a persuasivestructural argument for according the force of preemptive federal lawto some categories of customary international law, but fail to appreci-ate the full implications of their own argument.

III. THE MODERN POSITION, REDUX

Parts I and II showed that the modern position is more faithful tothe constitutional structure, the Founders’ intent, and pre-Erie doc-trine than is the revisionist position. But my defense of the modernposition so far has assumed that customary international law hasremained static over that period. This Part considers another prongof the revisionist argument: that the modern position should be aban-doned because the customary international law of today differs in rele-vant respects from the state-to-state portion of the law of nations asknown to the Founders. Revisionists note two important differences.First, customary international law today relies to a lesser extent onconsistent state practice. Second, modern customary internationallaw extensively addresses how a state may behave toward its ownnationals. I conclude that the first difference might warrant a gener-ally applicable limitation for the application of customary interna-tional law along the lines of the limitation articulated in Sabbatino andSosa, but it does not justify a wholesale rejection of the modern posi-tion. This limitation should also assuage concerns deriving from thesecond difference between historic and modern customary interna-tional law.

A. The New Ways of Making Customary International Law

The revisionists’ challenge to the federal status of customaryinternational law has usually been coupled with an attack on the“new” forms of international law.558 The traditional view was that cus-tomary international law was formed over a long period of time

558 See, e.g., Bradley & Goldsmith, Critique, supra note 9, at 855, 869–70; Young, Rsupra note 19, at 389. See generally Bradley & Goldsmith, Current Illegitimacy, supra note R9, at 319. R



through consistent state practice coupled with opinio juris, that is, theunderstanding that the practice was being undertaken out of a senseof legal obligation. Revisionists have noted that, in recent years, therequirements for creating customary international law have been con-siderably eased.559 The length of time necessary to form a new rulehas become shorter, to the point that today it is sometimes recognizedthat custom can form instantly.560 Relatedly, they note, the require-ment of consistent state practice has given way to an understandingthat custom can develop through statements or declarations of statesnot necessarily coinciding with state practice, such as their votes onnonbinding U.N. resolutions.561 As a result of the recognition ofthese new and easier ways of producing international law, the possibil-ity of claiming the existence of many new norms of customary interna-tional law has grown enormously, thus increasing significantly thepotential preemptive effect of this law.562

One possible response to this concern is to point out that thesedevelopments are controversial even within the international legalcommunity.563 However, this response merely brings to the fore theproblem of the indeterminacy of much customary internationallaw.564 The fact that international jurists cannot agree on what it takesto produce a rule of customary international law is hardly comforting.

To reassess the status of customary international law because ofchanges in its nature or content since the Founding is certainly legiti-mate. The domestic legal force of customary international law, unlikethat of treaties, is not specified in the constitutional text. The modernposition is based on an inference from the constitutional structure,and changes in the content or nature of such law might bear on the

559 As noted above, at the time of the Founding the law of nations was regarded ashaving, in part, a natural law foundation and not thought to be “made” at all. Seesupra note 369 and accompanying text. To that extent, the difference between the old Rand new customary international law is even starker.560 See RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW § 102 n.2 (1987) (recog-

nizing the concept of coastal state economic rights over the continental shelf as soquickly accepted after its announcement in 1945 that it became “instant custom[ ]”);Bin Cheng, Custom: The Future of General State Practice in a Divided World, in THE STRUC-

TURE AND PROCESS OF INTERNATIONAL LAW 513, 532 (R. St.J. Macdonald & Douglas M.Johnston eds., 1986).561 See, e.g., Bradley & Goldsmith, Critique, supra note 9, at 839. R

562 Revisionists make a point of describing the exorbitant claims that are occasion-ally made in the name of customary international law. See, e.g., id. at 841.563 See, e.g., Prosper Weil, Towards Relative Normativity in International Law?, 77 AM.

J. INT’L L. 413, 435 (1983).564 See generally, e.g., MARTTI KOSKENNIEMI, FROM APOLOGY TO UTOPIA (1989)

(describing such indeterminacy).



structural analysis. For example, such changes might reasonably leadto a reevaluation of the extent to which State violations of such lawwould compromise the federal government’s achievement of thenation’s foreign relations goals or the extent to which treating suchlaw as preemptive would improperly shift the balance of federal-Statepower.

The revisionists’ concerns about the indeterminacy of customaryinternational law and the easing of the requirements for recognizingnew norms of customary international law do seem germane to thestructural argument for regarding such law as preemptive. Thegreater the indeterminacy of the law, the broader the power of judgesto hold the laws of the States invalid. Although the indeterminacy ofthe norms does not necessarily result in judicial lawmaking,565 suchindeterminancy could in practice lead to broader preemption. Judi-cial deference to the executive branch on questions about the contentof such law might alleviate such concerns, but it would raise a concernabout circumvention of the bicameralism requirement. This latterconcern is exacerbated by the possible loosening of the requirementof state practice. The executive branch is the nation’s primary voiceon the international stage regarding the content of customary interna-tional law.566 According preemptive effect to law that is “made” byexecutive branch statements would appear to give the executivebranch a degree of lawmaking power that is at least in tension with theConstitution’s carefully wrought procedures for creating preemptivefederal law.567

But these problems do not afflict all of customary internationallaw. Many norms are well-established and valid under standards thatvirtually all international jurists would endorse.568 The fact that thereis disagreement at the margins, and that certain of the controversialtheories of international law formation are problematic in themselves,does not justify the denial of federal status to norms of customaryinternational law that are well settled under uncontroversial theories.

565 See supra text accompanying notes 73–74. R

566 See HENKIN, supra note 230, at 42–45. R

567 Cf. infra text accompanying notes 606–08 (arguing that States are required to Radhere to the executive branch’s interpretation of customary international law inincorporating as State law international legal norms that lack the force of supremefederal law).568 Professor Goldsmith has come close to denying that customary international

law exists at all. See JACK L. GOLDSMITH & ERIC A. POSNER, THE LIMITS OF INTERNA-

TIONAL LAW 42–43 (2005). But he has recognized that, for example, the immunity offoreign state officials exists and has the status of customary international law. SeeBradley & Goldsmith, supra note 199, at 16. R



A limitation along the lines of what the Sabbatino Court imposed forthe application of customary international law to invalidate the acts offoreign states within their own territory seems well tailored to addressthis problem. As noted, the Court held that norms that were disputedand touched sharply on national nerves could not be applied in thiscontext. The Court in Sosa applied a similar standard to identify thenorms that could support a federal common law action for damages.Although the Sabbatino limitation only applies to violations of custom-ary international law attributable to foreign states, and the similar lim-itations articulated in Sosa were only found applicable to actions fordamages, and may only apply when the acts of foreign states are beingchallenged, the concerns discussed in this section may justify abroader holding that only norms of customary international law thatsatisfy a heightened standard of clarity and acceptance count as pre-emptive federal law.569 Such a holding would be consistent with thestructural rationale for according customary international law the sta-tus of preemptive federal law. The structural argument is based onthe concern that violations will produce international friction andthus hinder the accomplishment of foreign relations goals. Denyingpreemptive force to debatable norms would not produce significantfriction, it might be claimed, precisely because the status of the normson the international plane is itself unclear.

Applying this heightened standard of clarity and acceptanceacross the board, however, might be problematic in a differentrespect. Such an approach would sometimes result in the applicationof a narrow rule in circumstances in which the Executive favors abroader rule.570 For example, the Executive may wish to enjoy abroad immunity in the courts of other states, but, if other states favora narrower rule and our courts applied the narrower rule (becausethe broader rule is disputed), the Executive would be on weak groundin claiming the broader immunity when the United States or its offi-cials are sued in the courts of other states. It may thus be preferablein some contexts for the courts to follow the Executive’s position if theExecutive favors a disputed rule. In Sabbatino, rule-of-law concerns led

569 This may be what Chief Judge Hand had in mind when he suggested that aState’s violation of a “well-established” principle of international law might raise afederal question. See supra text accompanying note 193. R570 A rule is “narrow” or “broad,” as I use these terms in the text, as measured

against the extent to which they restrict the rights of sovereign states. A narrow ruleleaves states with less freedom or discretion, whereas a broader rule leaves states withmore. Thus, a rule prohibiting the taking of the property of aliens unless accompa-nied by prompt, adequate, and effective compensation is narrower than a rule permit-ting such takings.



the Court to prefer a heightened standard, but there the Executivefavored a narrower rule than what the Court believed was universallyaccepted. Where the Executive favors according foreign states a nar-rower discretion than what well-accepted international law accords,following the Executive’s position in cases challenging the acts of for-eign states would produce rule-of-law problems and exacerbate inter-national friction. Where the Executive prefers a rule giving sovereignstates broader rights than what well-accepted international law accordsthem (as in the immunity example), then following the Executive incases seeking to apply the rule to foreign states would raise rule-of-lawconcerns but would not produce international friction. It may thus beappropriate for the courts to follow the Executive in the latter contextbut not the former.571

A heightened standard of clarity and breadth of acceptancewould reduce the range of norms of customary international law hav-ing preemptive effect, but, in contrast to the approach favored by revi-sionists, well-established norms of customary international law wouldcontinue to have preemptive effect. A noteworthy example is one thatwas implicated in the Sosa case but not considered by the SupremeCourt. Mr. Alvarez-Machain claimed that Mr. Sosa’s actions violatedthe rule of customary international law prohibiting the officials oragents of one nation from exercising police power on the territory ofanother nation without the latter nation’s consent, but the court ofappeals dismissed this claim. While acknowledging that “[f]ew princi-ples in international law are as deeply rooted as the general normprohibiting acts of sovereignty that offend the territorial integrity ofanother state,”572 the court held that the right belonged to Mexicoand thus Alvarez-Machain lacked standing to assert it.573 For reasonsthat I have addressed at length elsewhere, the court’s standing hold-

571 In cases challenging the acts of the States, a policy of following the Executive’spreferences where it favors a rule restricting the discretion of the States would notproduce international friction, but would raise greater federalism concerns, particu-larly in cases involving a State’s treatment of its own citizens. For this reason, it maybe appropriate for the courts to apply a heightened standard of clarity and breadth ofacceptance for norms restricting the discretion of the States even if the Executivefavors a disputed rule.572 Alvarez-Machain v. United States, 331 F.3d 604, 615 (9th Cir. 2003) (citing 1 L.

OPPENHEIM, OPPENHEIM’S INTERNATIONAL LAW § 119 (Sir Robert Jennings & SirArthur Watts eds., 9th ed. 1992)), rev’d on other grounds sub nom. Sosa v. Alvarez-Machain, 542 U.S. 692 (2004); see F.A. Mann, Reflections on the Prosecution of PersonsAbducted in Breach of International Law, in INTERNATIONAL LAW AT A TIME OF PERPLEXITY

407, 407 & n.2 (Yoram Dinstein & Mala Tabory eds., 1989)).573 Alvarez-Machain, 331 F.3d at 615–16.



ing was erroneous.574 The court’s reasoning would equally deny dip-lomats standing to assert diplomatic immunity, since such immunityis, as a matter of international law, a right of the states and not of theindividual.575 But the fact that the sending state may waive the diplo-mat’s immunity has never been thought to mean that the diplomatlacked standing to assert the immunity in the absence of waiver. Simi-larly, the fact that Mexico may consent to the exercise of police powerby U.S. agents on its soil should not deprive the victim of an unlawfulexercise of extraterritorial police power of standing to invoke the rele-vant rule of international law in our courts in the absence of consent.Indeed, the court of appeals’ analysis would have the paradoxicaleffect of restricting the range of domestically enforceable norms ofcustomary international law to a subset of such norms—those relatingto human rights—that the revisionists have expressed particular con-cerns about, and which for the most part did not even exist beforeErie, when all agree that customary international law was commonlyapplied by the courts.576

Because Alvarez-Machain prevailed in the court of appeals onother grounds and did not cross-petition for certiorari,577 the standingissue was not before the Supreme Court in Sosa. Had it consideredthe claim, the Court would presumably have reversed, just as itrecently reversed the court of appeals’ similar holding in United Statesv. Bond578 that an individual lacks standing to enforce Tenth Amend-ment limitations because that Amendment protects States, not individ-uals.579 Had the Court reversed the standing holding in Sosa, it wouldsurely have agreed with the court of appeals that this norm of custom-ary international law satisfied the heightened standard of clarity andbreadth of acceptance that it adopted in that case. The fact that Mr.Sosa was acting at the behest of an official of the executive branchraises the question whether there was a “controlling executive act”precluding the enforcement of the norm, an issue beyond the scopeof this Article.580 But there should be no question that, had Sosa been

574 See Vazquez, supra note 172, at 1086–97. R575 See Vienna Convention on Diplomatic Relations and Optional Protocol on Dis-

putes art. 32(1), opened for signature Apr. 18, 1961, 23 U.S.T. 3227, 3241, 500 U.N.T.S.95 (“The immunity from jurisdiction of diplomatic agents . . . may be waived by thesending State.”).576 See Vazquez, supra note 172, at 1094–95. R577 See Sosa, 542 U.S. at 699 (noting that the Ninth Circuit found for Alvarez-

Machain on both his Alien Tort Statute and Federal Tort Claims Act claims); Brief inOpposition to Petition for Writ of Certiorari, Sosa, 542 U.S. 692 (No. 03-339).578 581 F.3d 128 (3d Cir. 2009), rev’d, No. 09-1227 (U.S. June 16, 2011).579 Bond v. United States, No. 09-1227 (U.S. June 16, 2011).580 See supra text accompanying notes 223–32. R



acting at the behest of the State of California, or had he been an offi-cial of the State of California, his actions would have violated a normof international law meeting any conceivable standard of clarity andbreadth of acceptance, and would accordingly have been redressableunder the standard suggested herein. If the revisionist view wereadopted, on the other hand, California’s violation of this well-estab-lished principle of international law would not be judiciallyredressable.

In sum, revisionists have raised some valid concerns stemmingfrom the indeterminacy of some modern customary international lawand uncertainties about the process for forming new customary lawrules, but the solution they propose is overbroad. A better-tailoredresponse would simply limit the range of preemptive customary inter-national law to those norms that satisfy a heightened standard of clar-ity and breadth of acceptance—a standard that the Court has alreadyadopted in two narrow settings.581 In certain contexts, this height-ened standard may have to be coupled with an executive-deferenceapproach. But a wholesale rejection of the modern position isunwarranted.

B. The New Topics Addressed by Customary International Law

Some revisionists have also noted that the customary interna-tional law of today differs from the international law of the Founders’time in that it addresses how a nation treats its own nationals.582

International law has always addressed how nations may behavetoward individuals. But, before the Second World War, internationallaw addressed individuals almost exclusively by limiting how onenation could treat the citizens of another nation. After World War II,with the broad recognition of an international law of human rights,international law came to address more extensively the behavior ofnations toward their own citizens.

But the international law of human rights implicates the struc-tural reasons for according preemptive force to customary interna-tional law no less than the older topics covered by customary

581 In a recent book chapter, Kenneth Randall and Chimene Keitner describe thesubset of customary international law that the Court found enforceable in Sabbatinoand Sosa as “supernorms” and they suggest that the Court has already adopted abroader rule limiting enforcement of customary international law to this subset. SeeKenneth C. Randall & Chimene I. Keitner, Sabbatino, Sosa, and “Supernorms,” inLOOKING TO THE FUTURE 559 (Mahnoush H. Arsanjani et al. eds., 2011). For thereasons discussed above, I do not think it is correct to say that the Court has alreadyembraced this as a general limitation.582 See, e.g., Young, supra note 19, at 368, 420. R



international law do. Even though human rights law is based on theidea that individuals have certain rights by virtue of their humanity,one consequence of viewing these norms as part of international law isthe recognition that one nation’s violations of these norms is the con-cern of the world community. A nation’s obligations under the inter-national law of human rights are obligations toward other states, not justtoward individuals.583 Thus, the United States’ violation of thesenorms is as likely to produce international friction—and thus to com-plicate the nation’s pursuit of foreign relations goals—as its violationof other norms of customary international law.

The growth of human rights law does implicate a different sort ofconcern relevant to the debate about the modern position. Suchnorms are more likely to relate to areas that have traditionally beenregulated by the States, and in some respects may relate to mattersthat have been thought to be constitutionally reserved to the States.According preemptive effect to international human rights normsthus risks a far broader preemption of State law than had been tradi-tionally associated with customary international law.584 It is likely,however, that this concern would be adequately addressed by limita-tions along the lines of those articulated in Sabbatino and Sosa, ifextended beyond the particular contexts of those cases. The humanrights norms that most concern the revisionists are actually highly con-troversial within the international legal community. For example,revisionists frequently invoke the claimed norm prohibiting the impo-sition of the death penalty for crimes committed by juveniles.585 Asnoted above, if this is a norm of customary international law, it is likelythat the United States is exempt from it as a persistent objector.586 Ifonly norms meeting a high threshold of clarity and acceptance were

583 See RENE PROVOST, INTERNATIONAL HUMAN RIGHTS AND HUMANITARIAN LAW 222(2002) (“Human rights as a whole . . . are generally considered to generate erga omnesor erga omnes partes obligations for states.”); Barcelona Traction, Light & Power Co.(Belgium v. Spain), 1970 I.C.J. 3, ¶¶ 33–34 (Feb. 5, 1970) (“By their very nature [theobligations of a State towards the international community as a whole] are the con-cern of all States. In view of the importance of the rights involved, all States can beheld to have a legal interest in their protection; they are obligations erga omnes . . . .Such obligations derive, for example, in contemporary international law, from theoutlawing of acts of aggression, and of genocide, as also from the principles and rulesconcerning the basic rights of the human person, including protection from slaveryand racial discrimination.”). For this reason, the international law of human rightscorresponds to the state-to-state branch of the Founding era’s law of nations, not thelaw merchant or general commercial law.584 See Young, supra note 19, at 420–23 (advancing this argument forcefully). R

585 See id. at 474.586 See supra note 426 and accompanying text. R



deemed to have preemptive force, it is very likely that the humanrights norms that would preempt State law would largely duplicateprohibitions imposed on the States by the Constitution. (Indeed, thatis currently the case with respect to the claimed norm of customaryinternational law prohibiting the juvenile death penalty.587)

If, contrary to my argument, the customary international law ofhuman rights were regarded as too different from the more tradi-tional kind of customary international law because of its greater over-lap with subjects traditionally reserved to the States, or for some otherreason, the solution would be to deny preemptive force to this subsetof customary international law, not to reject the modern position alto-gether. Such an approach would continue to accord preemptive forceto the more traditional sorts of customary international law, such asthe norm that was erroneously dismissed from the Sosa case before itreached the Supreme Court. The Court’s analysis in Sosa of theclaimed human rights violation before it suggests that it agrees withmy conclusion that a heightened standard of clarity and breadth ofacceptance suffices to protect the relevant constitutional interests andthat a categorical exclusion of human rights norms from the range ofpreemptive federal law is unwarranted.

IV. STATE INCORPORATION OF CUSTOMARY INTERNATIONAL LAW

Revisionists maintain that customary international law has theforce of law in this country only to the extent that it has been affirma-tively incorporated as such. It has the force of federal law if incorpo-rated by the federal political branches, ordinarily through federalstatute or treaty. In the absence of federal incorporation, it has theforce of State law if incorporated by the States. This Article so far hasaddressed only the first part of that claim and concluded that the bet-ter answer from the perspective of constitutional history and structureis that customary international law has the status of preemptive fed-eral law, with some exceptions for suits challenging the conduct offoreign states and some federal officials. I have argued further thatthe revisionists’ arguments based on the differences between the cus-tomary international law of today and the state-to-state branch of thelaw of nations as known to the Founders can be adequately addressedby limiting the scope of preemptive customary international law tonorms satisfying requirements along the lines of those articulated inSabbatino and Sosa.

587 See Roper v. Simmons, 543 U.S. 551, 575–79 (2005).



I now consider the second part of the revisionist claim: that, tothe extent a norm of customary international law lacks the force ofpreemptive federal law, the States have the power to incorporate it asState law. Contrary to the apparent assumption of the revisionists,State incorporation of customary international law norms that do nothave the force of preemptive federal law could potentially cause astructural problem of its own. The scope of this problem would bebroader if, as revisionists argue, customary international law does notgenerally have the force of preemptive federal law. If only disputed orunclear norms lack such status, the scope of the problem would benarrower, but the problem would nevertheless exist. This Part identi-fies the structural problem and argues that the solution is either toprohibit States from incorporating customary international law asState law or to recognize the jurisdiction of the Supreme Court toreview the State courts’ interpretations of customary international laweven when such law is applicable to the case only as State law. I con-clude that the latter solution is preferable and consistent with ArticleIII.

It is clear that federal law limits the power of States to incorporatecustomary international law as State law at least in some contexts. Forexample, Sabbatino holds that certain norms of customary interna-tional law may not be applied as a basis for invalidating the acts of aforeign state performed within its own territory. The reason is thatthe failure to give effect to the act of state would hinder the politicalbranches’ pursuit of the nation’s foreign relations interests. TheCourt in Sabbatino explicitly held that the State courts, no less than thefederal courts, were prohibited from applying these norms in this con-text.588 The State courts’ failure to give effect to such acts of statewould be just as likely to hinder the federal government’s conduct offoreign relations. Thus, the constitutional structure disables theStates from incorporating certain norms of customary internationallaw as State law against foreign states acting within their own territory.

The Court in Sosa did not directly address whether the Stateswere permitted to recognize a right of action for damages to enforcenorms of international law not meeting the heightened standard ofclarity and breadth of acceptance that it articulated. The Court’s anal-ysis, however, strongly suggests a negative answer, at least with respectto certain defendants. In limiting the federal right of action to thissubset of customary international law norms, the Court in Sosaexpressed concern about the foreign relations sensitivities ofextending the right of action to less-well-established norms and the

588 See Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398, 427 (1964).



potential interference with the “discretion of the Legislative and Exec-utive Branches in managing foreign affairs” that could result.589

These same structural concerns would result from the recognitionand enforcement of State law rights of action. Thus, it would appearthat a State Alien Tort Claims Act with less stringent requirementsthan contemplated by Sosa would be preempted by the same struc-tural constitutional concerns that led the Court to limit the federalright of action.

It is true that the Court in Sosa expressed concerns about foreignrelations sensitivities in discussing claims alleging the violation ofinternational law by foreign officials abroad (a situation very similar tothat in Sabbatino).590 Its reasoning may not apply beyond a Filartiga-type situation, where customary international law is sought to beapplied to the acts of foreign officials. But, to the extent that a Stateright of action were made applicable to federal officials, the interfer-ence with federal functions would appear to be even more direct andhence more problematic. As discussed above, customary internationallaw as federal law does not bind certain federal officials in certain con-texts. Surely, the extent to which federal officials are bound is purelya matter of federal law.

That leaves a State’s incorporation of customary international lawto bind its own officials. On the surface, it may seem that a State’sincorporation of customary international law to bind itself poses noproblem from a structural constitutional standpoint.591 But a closerexamination reveals a potential problem resulting from the processunder international law for creating new rules of customary interna-tional law. The traditional rule has been that customary internationallaw is formed through consistent state practice performed out of asense of legal obligation. As noted by revisionists, newer theories havegained some recognition under which customary international lawmay be formed through the public statements of states indicatingtheir views about what is required by customary international law. IfStates were free to incorporate customary international law as Statelaw in circumstances where such law lacks the status of federal law,even if only to bind their own officials, then State courts would bemaking pronouncements about what international law requires in cer-tain contexts. Such decisions would constitute state practice and evi-

589 See Sosa v. Alvarez-Machain, 542 U.S. 692, 727 (2004).590 See supra text accompanying notes 254–55. R591 Cf. Fletcher, supra note 19, at 670 (“If a [S]tate court decides that the death R

penalty should be forbidden in prosecutions brought under [S]tate law, such a deci-sion is entirely that [S]tate’s business [regardless of whether it is made on the basis ofcustomary international law].”).



dence of opinio juris under the traditional approaches to theformation of customary international law. Additionally, such deci-sions are a “subsidiary means for the determination of rules of [cus-tomary international] law.”592 Under the newer approaches, Statecourt decisions would be even more likely to be relevant to the forma-tion of new rules of customary international law. Such State courtpronouncements about what international law requires would thuscontribute to the formation of new rules that could, in turn, bind thefederal government.

It is well accepted that the decisions of domestic courts count asstate practice for the purpose of forming customary international law.As explained by the International Law Association (ILA) in its State-ment of Principles Applicable to the Formation of General Customary Interna-tional Law, “[d]omestic courts . . . are organs of the State, and theirdecisions should also be treated as part of the practice of the State.For example, a determination that international law does or does notrequire State immunity to be accorded in a particular case, or theextraterritorial application of a domestic law.”593 State courts are asmuch “domestic courts” for international law purposes as federalcourts. Their actions are as attributable to the nation for purposes ofstate responsibility as those of federal courts.594 According to theILA’s Statement of Principles, “[t]he activities of provinces within a fed-eration, or of other subordinate local authorities” may constitute statepractice “if the entity concerned acts with the authority of the (fed-eral) State, or if the latter adopts its acts.”595 “A [federal] State’s fail-ure to prevent the conduct in question can amount, for presentpurposes, to tacit adoption . . . (whether or not the [national govern-ment] is in a position, under its constitutional law, to change the rulesin question).”596 Thus, if the U.S. Constitution were construed to per-mit the States to incorporate customary international law as State law,then State pronouncements about what customary international lawrequires would count as state practice. The Constitution would, insuch circumstances, be understood to delegate to the States the powerto take positions on what customary international law requires inthese contexts, and the States’ decisions would be considered to havebeen tacitly adopted by the federal government, even if the federalcourts had no power to review those decisions.

592 Statute of the International Court of Justice art. 38(1)(d), June 26, 1945, 59Stat. 1055, 1060, 33 U.N.T.S. 993.593 ILA STATEMENT OF PRINCIPLES, supra note 71, at 18 (footnote omitted). R594 See Responsibility of States, supra note 94, art. 4 & cmt. 6. R595 ILA STATEMENT OF PRINCIPLES, supra note 71, at 16–17. R596 Id. at 17.



Such decisions could thus contribute to the evolution of newrules of international law that would bind the federal governmenteven if it disagreed with them. If a State court ruled that customaryinternational law imposes a particular rule, then the State court’s deci-sion, in combination with the practice of other states, could wind upcrystallizing into a new rule of international law. The likelihood ofthat happening is perhaps remote, since it would require action bynumerous other nations in conjunction with that of the State court,and the federal government could make statements disagreeing withthe State court decision, thus diluting the impact of the State courtdecision internationally. But if the rule did come into being, the Statecourt’s decisions in conflict with the federal view might compromisethe United States’ ability to claim persistent objector status.

A second scenario is more likely to occur: the State court mightdecide that a norm of customary international law does not exist eventhough the federal government believes that it does or should exist.In this way, the State court judgment could contribute to the eviscera-tion of an existing norm of international law. Since it takes less statepractice to vitiate a rule of international law than to establish one,597

the State’s conduct is more likely to have the feared effect. Addition-ally, a State court’s decision could hinder the federal government’sefforts to establish a new rule of international law. Although the fed-eral Executive’s position will usually be accorded more weight in thiscontext, “the internal uniformity or consistency which is needed for aState’s practice to count towards the formation of a customary rulemay . . . be prejudiced.”598

The ability of States to incorporate customary international law asState law would thus interfere with the federal government’s conductof foreign relations by hindering its ability to establish norms that itchampions and, potentially, by contributing toward the establishmentof norms that it does not favor, and which could ultimately bind it. Inthe context of dormant foreign affairs preemption, the “one-voice”argument has often been criticized,599 and the Court itself has dis-tanced itself from the argument in related contexts.600 But the “one-voice” argument has particular resonance with respect to customaryinternational law because of the way customary international lawevolves. State court decisions regarding the requirements of custom-

597 See Carlos M. Vazquez, Withdrawing from International Custom: Terrible Food, SmallPortions, 120 YALE L.J. ONLINE 269, 279–80 (2011), http://www.yalelawjournal.org/images/pdfs/955.pdf.598 ILA STATEMENT OF PRINCIPLES, supra note 71, at 18. R599 See, e.g., Ramsey, supra note 17, at 561–63; Young, supra note 19, at 447–50. R600 See Barclays Bank v. Franchise Tax Bd., 512 U.S. 298, 320–24 (1994).



ary international law may dilute the federal government’s voice whenit speaks in favor of a particular norm, and may even contradict andeventually bind the federal government to norms it opposes.

There are two possible solutions to this problem. The first wouldbe to deny the States the power to incorporate customary interna-tional law as State law in a manner that requires State judicial oradministrative officials to issue pronouncements about what custom-ary international law requires. The second, more limited, approachwould permit such incorporation, but would recognize the power ofthe Supreme Court to review the decisions of State courts or adminis-trative agencies about what customary international law requires, evenwhen such law is relevant to the case only because it has been incorpo-rated into State law.

Under the first approach, the Supreme Court would simply strikedown any State law incorporating customary international law in themanner described. The standard for finding norms of customaryinternational law to be preemptive federal law would thus function asa ceiling as well as a floor. In other words, States would be required tocomply with and enforce norms satisfying the standard and prohibitedfrom incorporating norms that do not satisfy it. To be clear, the Stateswould not be precluded from legislating out of a belief that legislationis necessary to carry out international law. Thus, they would be free toprohibit the execution of minors out of a belief that customary inter-national law prohibits such executions, even if their belief is mistaken.The structural problem arises when the State courts make pronounce-ments about the content of international law. Thus, a statute prohib-iting penalties that would violate customary international law wouldbe invalid insofar as it is construed to apply to norms that do not meetwhatever standard is adopted for giving preemptive effect to custom-ary international law norms, whereas a statute that prohibited the exe-cution of minors would not be invalid, even if it was motivated by amistaken belief that such executions violate customary internationallaw. Because only public acts count as state practice for purposes ofcustomary international law formation,601 a State legislature’s motivefor enacting a law would not cause the same structural problem as aState court’s decision that particular conduct violates customary inter-national law.602

601 See ILA STATEMENT OF PRINCIPLES, supra note 71, at 15 (“Acts do not count as Rpractice if they are not public.”).602 The structural problem would perhaps arise if the legislature’s motive were

memorialized in a formal document, such as a legislative report or the statute’s pre-amble, but only in diluted form.



The second approach would address the structural problem bypermitting the Supreme Court to review the State courts’ interpreta-tions of customary international law. This approach seems preferableto the first. The structural problem arises only if the State courts haveinterpreted international law erroneously, but the State courts maywell have gotten the international law right. To be sure, the SupremeCourt may lack the expertise to decide disputed or unsettled ques-tions of customary international law. But, in this context, the Court’srole would be simply to follow the executive branch’s interpretation ofcustomary international law. The particular structural problem arisesbecause the State’s action will contribute to the formation of a normthat could come to bind the federal government. Because the Presi-dent speaks for the nation at the international plane concerning thecontent of customary international law,603 the problem disappears ifthe Executive agrees with the State court’s interpretation of whatinternational law requires.

A possible obstacle to Supreme Court review of State court deci-sions in this context would be Article III of the Constitution. TheSupreme Court clearly has jurisdiction to review the State courts’interpretations of federal law, but in this context, by hypothesis, cus-tomary international law is being applied as State law. True, theSupreme Court can review federal issues embedded in State lawcauses of action.604 It is thus clear that the Supreme Court can reviewState court decisions even if the remedy being sought in the case isavailable only by virtue of State law. But in such cases there is nodoubt that the issue being reviewed is one of federal law. Here, byhypothesis, the relevant norm lacks the status of preemptive federallaw.

But, once we take account of the structural constitutional basisfor Supreme Court review in this context, we see that the Article IIIproblem is illusory. Assume that Congress passed a statute preempt-ing State laws incorporating an erroneous interpretation of customaryinternational law. Clearly, Supreme Court review of the State court’sinterpretation of customary international law would be consistent withArticle III under such circumstances. The same would be true if thestatute preempted State laws incorporating an interpretation of inter-national law that differed from that espoused by the executive branch.

603 See HENKIN, supra note 230, at 43–44. R

604 See Grable & Sons Metal Prods., Inc. v. Darue Eng’g & Mfg., 545 U.S. 308(2005) (stating that cases arising under State law in which a federal issue is embeddedmay “arise under” federal law for purposes of the original jurisdiction of the districtcourts).



My claim here is that State court interpretations of disputed or contro-versial norms of customary international law that differ from that ofthe executive branch regarding the content of such law are problem-atic from the perspective of the constitutional structure. In otherwords, the Constitution itself accomplishes what the federal statuteposited above would accomplish.605 The case thus “arises under” theConstitution, which establishes that the federal Executive speaks forthe nation regarding the content of customary international law inthis context

I argued in Part II that a rule-of-law problem would arise if thecourts simply followed the Executive’s views about the content of cus-tomary international law.606 Because of its concern that the contentof the law enforced by the courts would change with each successiveadministration, and that the courts would then be reduced to theExecutive’s errand boys, the Court in the act-of-state context pre-ferred to declare whole categories of customary international law judi-cially unenforceable in the face of certain acts of foreign states.607 Forthe same reason, I argued above that a broader exclusion of disputedor unclear norms of customary international law from the range ofpreemptive federal law might be preferable to a regime in which thecourts simply followed the Executive on a case-by-case basis.608 In thecontext of State incorporation of customary international law normsthat lack the status of preemptive federal law, however, it is proper totreat the Executive’s views as binding. If my analysis in Part III werefollowed, then, by hypothesis, the problem under discussion wouldarise only with respect to disputed or unsettled norms of customaryinternational law. The structural constitutional concern is that aState’s pronouncement regarding the content of customary interna-tional law will inhibit the Executive’s efforts towards future crystalliza-tion of a norm or will contribute to the crystallization of a norm thatthe Executive opposes. Because the Constitution allocates to theExecutive the power to take positions on customary international lawfor the purpose of contributing to the possible future crystallization ofthe norm, it is appropriate to treat the Executive’s views as binding inthis context. Additionally, the concerns expressed in Part III aboutjudicial law-making are inapplicable in this context, as the courts areinvolved at all only because the States have made international law

605 True, the State statutes we are positing do not expressly incorporate any partic-ular interpretation of international law, but the statutes as construed by the Statecourts do.606 See supra text accompanying note 545. R607 See supra note 214. R608 See supra text accompanying note 545. R



relevant by incorporating it as State law. The states will have “made”the law, and the federal courts would only be enforcing a constitu-tional limitation on their ability to do so. (As noted, judicial review inthis context may be conceived as merely the enforcement of a consti-tutional rule prohibiting States from incorporating into State law aninterpretation of a disputed or controversial norm of customary inter-national law that differs from that of the Executive Branch.) Undersuch circumstances, it seems unproblematic for the courts to followthe Executive.

* * * *

In sum, there is a potential structural problem with permittingthe States to incorporate as State law norms of customary interna-tional law that lack the force of preemptive federal law: State pro-nouncements about what customary international law does and doesnot require may, over time and in conjunction with the statementsand practice of other states, wind up undermining norms that the fed-eral government favors, or establishing norms that the federal govern-ment does not endorse and which could eventually bind the federalgovernment. The structural problem can be cured, however, by rec-ognizing that, as a matter of federal constitutional law, States lack thepower to incorporate interpretations of disputed or evolving rules ofcustomary international law that differ from those espoused by thefederal government. The theory would be that the Constitution allo-cates to the federal government the power to speak for the nationconcerning disputed or evolving norms of customary internationallaw, and state court pronouncements regarding the content of suchlaw conflict with this constitutional allocation of power to the extentthat they differ from that of the federal government. If so, then theSupreme Court would have jurisdiction to review any state court deci-sion interpreting such norms of customary international law to ensurethat they comport with the federal government’s views, even if thenorm is relevant to the case solely because of a State has incorporatedit as State law.

CONCLUSION

This Article has shown that the modern position concerning thestatus of customary international law is more consistent with constitu-tional structure, original intent, and pre- and post-Erie doctrine thanthe revisionist position. It has shown further that an intermediate sta-tus for customary international law is untenable, and that all of thespecific intermediate proposals that have been advanced save that of



Bellia and Clark are highly problematic. The analysis of Bellia andClark, under which the preemptive force of some norms of customaryinternational law results from an inference from the constitutionalstructure, is basically sound but, rather than being an intermediateposition, it supports the modern position.

The revisionists have raised valid concerns about the differencesbetween the relevant portion of the law of nations as known to theFounders and the customary international law of today, but these con-cerns justify a limitation of the modern position, not its rejection. Theconcerns would be met if the range of norms of customary interna-tional law having preemptive effect were restricted to those meeting aheightened standard of clarity and breadth of acceptance.

The revisionist position is also potentially problematic insofar asit would permit the States to incorporate as State law norms of custom-ary international law that lack the status of preemptive federal law.Such incorporation poses a structural problem to the extent that theState courts’ pronouncements about the requirements of interna-tional law differ from those of the federal Executive. This problemcan be addressed either by disabling the States from incorporatingsuch norms as State law in a manner that produces State judicial oradministrative pronouncements about what this law requires or by rec-ognizing the Supreme Court’s power to review the State courts’ inter-pretations of customary international law even when its applicabilityto the case derives entirely from State law incorporation. The lattersolution is preferable and consistent with Article III.

Customary International Law as U.S. Law: A Critique of the

Documents

Transcript of Customary International Law as U.S. Law: A Critique of the